Skip to content

利用 AI 大模型和自动化工作流,根据主题或关键词一键生成高清短视频。Generate HD short videos from a topic or keyword with an automated AI workflow.

feat(material): add OFox multi-model video provider

celeste1900ba49f3426da4
diff --git a/README-en.md b/README-en.md
index 23175407..31c56267 100644
--- a/README-en.md
+++ b/README-en.md
@@ -141,6 +141,7 @@ Thanks to [Kimi](https://platform.kimi.ai?track_id=track-f6b0a640d35c41deb03b247
- [x] Supports your own **local assets** and free-to-use HD footage from **Pexels**, **Pixabay**, and **Coverr**
- [x] Supports **AI-generated footage**: [WaveSpeed AI](https://wavespeed.ai) text-to-video models (Seedance by default) create brand-new visuals from your script keywords instead of relying on stock libraries
- [x] Supports native **Volcano Engine Ark Seedance** text-to-video generation with configurable model/Endpoint ID, bounded polling, and paid-task confirmation
+- [x] Supports native **OFox** multi-model text-to-video generation (Seedance, Wan, and more behind one API key) with bounded polling and paid-task confirmation
- [x] Supports leading model providers including **Kimi / Moonshot AI**, **OpenAI**, **Anthropic Claude**, **Google Gemini**, **DeepSeek**, **Alibaba Cloud Qwen**, **Microsoft Azure OpenAI**, **ByteDance VolcEngine Ark**, **xAI Grok**, **MiniMax**, and **Xiaomi MiMo**, plus unified gateways, aggregators, and local runtimes such as **Cloudflare AI Gateway**, **Alibaba ModelScope**, **AIHubMix**, **AIML API**, **EvoLink**, **Ollama**, **OneAPI**, **LiteLLM**, **Groq**, and **Pollinations AI**
- [x] Supports one-click **cross-platform publishing** to **TikTok**, **Instagram**, and **YouTube Shorts** after video generation
- [x] Supports **exporting and importing generation settings** as a preset file, and backing up and restoring every **API key** from the settings dialog
@@ -483,6 +484,24 @@ The first integration supports text-to-video only. Every submitted clip is a pai
</details>
+<details>
+<summary>How do I use the OFox multi-model text-to-video provider?</summary>
+
+Create an [OFox API key](https://ofox.ai), then configure under `[app]`:
+
+```toml
+[app]
+ofox_api_key = "your-ofox-api-key"
+ofox_text_to_video_model = "bytedance/seedance-2.0-fast"
+ofox_base_url = "https://api.ofox.ai/v1"
+```
+
+When the dedicated config value is empty, the `OFOX_API_KEY` environment variable is used. Select **OFox AI Video** as the video source and explicitly confirm paid generation before starting. CLI users must add `--confirm-ofox-charge`. The model can be any text-to-video model from the OFox catalog (`GET /v1/models`), e.g. `alibaba/wan-2.7`; each model has its own supported duration range and resolutions, and out-of-range values are rejected by the API with a clear error before any task is created.
+
+The first version supports text-to-video only. Each clip creates one asynchronous paid task; the program only polls the same task id, stops submitting new tasks whenever a paid task state is unknown, and generates just enough footage to cover the voiceover duration.
+
+</details>
+
<details>
<summary>RuntimeError: No ffmpeg exe could be found</summary>
diff --git a/README-ja.md b/README-ja.md
index 558cf1f0..5a0fd873 100644
--- a/README-ja.md
+++ b/README-ja.md
@@ -139,6 +139,7 @@
- [x] ランダムまたはカスタムの **BGM** に対応し、音量も調整できます
- [x] 手持ちの**ローカル素材**に加え、**Pexels**、**Pixabay**、**Coverr** の無料で使える高画質素材に対応
- [x] ネイティブの **Volcano Engine Ark Seedance** テキスト動画生成に対応し、モデル/Endpoint ID、制限付きポーリング、有料タスク確認を設定できます
+- [x] ネイティブの **OFox** マルチモデルテキスト動画生成に対応(1 つの API キーで Seedance や Wan などを利用可能)。制限付きポーリングと有料タスク確認も同様に提供します
- [x] **Kimi / Moonshot AI**、**OpenAI**、**Google Gemini**、**DeepSeek**、**Alibaba Cloud Qwen**、**Microsoft Azure OpenAI**、**ByteDance VolcEngine Ark**、**xAI Grok**、**MiniMax**、**Xiaomi MiMo** といった主要なモデルプロバイダーに加え、**Cloudflare AI Gateway**、**Alibaba ModelScope**、**AIHubMix**、**AIML API**、**EvoLink**、**Ollama**、**OneAPI**、**LiteLLM**、**Groq**、**Pollinations AI** などの統合ゲートウェイ、アグリゲーター、ローカルランタイムに対応
- [x] 動画生成後、**TikTok**、**Instagram**、**YouTube Shorts** へワンクリックで**クロスプラットフォーム投稿**が可能
- [x] 生成設定をプリセットファイルとして**エクスポート/インポート**でき、設定画面から **API キー**のバックアップと復元も可能
@@ -461,6 +462,24 @@ Seedance 専用設定が空の場合は、`VOLCENGINE_ARK_API_KEY` 環境変数
</details>
+<details>
+<summary>OFox のマルチモデルテキスト動画素材ソースを使うには?</summary>
+
+[OFox API キー](https://ofox.ai)を作成し、`[app]` 配下に設定します:
+
+```toml
+[app]
+ofox_api_key = "your-ofox-api-key"
+ofox_text_to_video_model = "bytedance/seedance-2.0-fast"
+ofox_base_url = "https://api.ofox.ai/v1"
+```
+
+専用設定が空の場合は、`OFOX_API_KEY` 環境変数を使用します。動画ソースで **OFox AI Video** を選び、生成前に有料タスクを明示的に確認してください。CLI では `--confirm-ofox-charge` も必要です。モデルは OFox カタログ(`GET /v1/models`)の任意のテキスト動画モデル(例:`alibaba/wan-2.7`)に変更できます。モデルごとに対応する動画長と解像度が異なり、範囲外の値はタスク作成前に API が明確なエラーで拒否します。
+
+初版はテキスト動画のみ対応です。各クリップは非同期の有料タスクを 1 つ作成します。プログラムは同じタスク ID のみをポーリングし、状態が不明な場合は新規タスクの送信を停止し、ナレーションの長さを満たす分だけ生成します。
+
+</details>
+
<details>
<summary>RuntimeError: No ffmpeg exe could be found</summary>
diff --git a/README.md b/README.md
index 2ea2c965..91097bd7 100644
--- a/README.md
+++ b/README.md
@@ -151,6 +151,7 @@
- [x] 支持使用自己的 **本地素材**,也可从 **Pexels**、**Pixabay** 和 **Coverr** 获取可免费使用的高清素材
- [x] 支持 **AI 生成素材**:接入 [WaveSpeed AI](https://wavespeed.ai) 文生视频模型(默认 Seedance),按脚本关键词直接生成全新画面,不再受限于库存素材
- [x] 原生支持 **火山引擎方舟 Seedance** 文生视频,可配置模型或推理接入点 ID,并提供有限轮询与付费任务确认
+- [x] 原生支持 **OFox** 多模型文生视频(一个 Key 可调 Seedance、Wan 等模型),同样提供有限轮询与付费任务确认
- [x] 支持 **Kimi / Moonshot AI**、**OpenAI**、**Anthropic Claude**、**Google Gemini**、**DeepSeek**、**阿里云通义千问**、**Microsoft Azure OpenAI**、**火山引擎方舟**、**xAI Grok**、**MiniMax**、**小米 MiMo** 等主流模型服务,并兼容 **Cloudflare AI Gateway**、**魔搭 ModelScope**、**AIHubMix**、**AIML API**、**EvoLink**、**Ollama**、**OneAPI**、**LiteLLM**、**Groq**、**Pollinations AI** 等统一网关、聚合平台和本地运行环境
- [x] 支持一键 **跨平台发布**,生成完成后可自动上传至 **TikTok**、**Instagram** 和 **YouTube Shorts**
- [x] 支持将生成设置**导出和导入**为预设文件,并在设置弹窗中备份和恢复所有 **API Key**
@@ -478,6 +479,24 @@ Seedance 专用配置为空时,会依次使用 `VOLCENGINE_ARK_API_KEY` 环境
</details>
+<details>
+<summary>如何使用 OFox 多模型文生视频素材源?</summary>
+
+创建 [OFox API Key](https://ofox.ai),然后在 `[app]` 下配置:
+
+```toml
+[app]
+ofox_api_key = "your-ofox-api-key"
+ofox_text_to_video_model = "bytedance/seedance-2.0-fast"
+ofox_base_url = "https://api.ofox.ai/v1"
+```
+
+专用配置为空时,会使用 `OFOX_API_KEY` 环境变量。在视频源中选择"OFox AI 视频",并在生成前明确确认付费任务;CLI 还需要传入 `--confirm-ofox-charge`。模型可换成 OFox 目录(`GET /v1/models`)中任意文生视频模型(如 `alibaba/wan-2.7`),不同模型支持的时长区间与分辨率不同,超出范围会在创建任务前被 API 明确拒绝。
+
+首版仅支持文生视频。每个片段都会创建一个异步付费任务;程序只轮询同一个任务 ID,状态不明时停止继续下单,并仅生成足够覆盖配音时长的素材。
+
+</details>
+
<details>
<summary>RuntimeError: No ffmpeg exe could be found</summary>
diff --git a/app/services/material.py b/app/services/material.py
index 818ceb3b..a2a24568 100644
--- a/app/services/material.py
+++ b/app/services/material.py
@@ -17,7 +17,13 @@ from PIL import Image, UnidentifiedImageError
from app.config import config
from app.models.schema import MaterialInfo, VideoAspect, VideoConcatMode
-from app.services import material_cache, task_artifacts, video, volcengine_seedance
+from app.services import (
+ material_cache,
+ ofox,
+ task_artifacts,
+ video,
+ volcengine_seedance,
+)
from app.utils import utils
# Thread-safe counter for API key rotation
@@ -1710,6 +1716,18 @@ def download_videos(
max_clip_duration=max_clip_duration,
material_directory=material_directory,
)
+ if source == "ofox":
+ # 与 WaveSpeed/方舟相同的按需付费语义:OFox 网关的 /v1/videos 会创建
+ # 异步付费任务,必须逐段生成、凑够所需时长立即停止;产物地址是会过
+ # 期的临时直链,也不参与 24 小时搜索缓存。
+ return _download_videos_ofox_on_demand(
+ task_id=task_id,
+ search_terms=search_terms,
+ video_aspect=video_aspect,
+ audio_duration=audio_duration,
+ max_clip_duration=max_clip_duration,
+ material_directory=material_directory,
+ )
if source == "openai_image":
# 与 WaveSpeed 相同的按需付费语义:文生图按张计费,逐段生成、凑够
# 所需时长立即停止。生成结果是一次性的本地图片文件,也不参与 24
@@ -1990,6 +2008,111 @@ def _download_videos_seedance_on_demand(
return video_paths
+def _download_videos_ofox_on_demand(
+ *,
+ task_id: str,
+ search_terms: List[str],
+ video_aspect: VideoAspect,
+ audio_duration: float,
+ max_clip_duration: int,
+ material_directory: str,
+) -> List[str]:
+ """顺序生成 OFox 素材,覆盖配音时长后立即停止付费下单。"""
+ video_paths: List[str] = []
+ material_sources: list[dict[str, Any]] = []
+
+ # 付费生成循环必须先验证控制循环次数的两个时长。NaN/Infinity 会让
+ # ``total_duration >= audio_duration`` 永远不成立,而非正片段时长会让
+ # 累计值无法增长,两者都可能为全部关键词创建无用的付费任务。
+ try:
+ required_duration = float(audio_duration)
+ except (TypeError, ValueError) as exc:
+ raise ofox.OFoxError("OFox audio duration must be a finite number") from exc
+ if not math.isfinite(required_duration):
+ raise ofox.OFoxError("OFox audio duration must be a finite number")
+ if required_duration <= 0:
+ logger.warning(
+ "skip OFox paid generation because required audio duration is "
+ f"not positive: duration={required_duration}"
+ )
+ _persist_material_sources(task_id, material_sources)
+ return video_paths
+
+ try:
+ clip_duration = int(max_clip_duration)
+ except (TypeError, ValueError, OverflowError) as exc:
+ raise ofox.OFoxError("OFox clip duration must be a positive integer") from exc
+ if clip_duration <= 0:
+ raise ofox.OFoxError("OFox clip duration must be a positive integer")
+
+ total_duration = 0.0
+ for search_term in search_terms:
+ try:
+ video_items = ofox.generate_videos(
+ search_term=search_term,
+ minimum_duration=clip_duration,
+ video_aspect=video_aspect,
+ )
+ except ofox.OFoxUnconfirmedTaskError as exc:
+ # 远端付费任务仍可能成功。立即停止继续下单,并保留任务 ID,方便
+ # 用户随后在 OFox 控制台确认或找回结果。
+ logger.error(
+ "stop submitting new OFox tasks because the last paid task "
+ f"is unconfirmed: task_id={exc.task_id or 'unknown'}, detail={exc}"
+ )
+ _persist_material_sources(task_id, material_sources)
+ raise
+ except ofox.OFoxError as exc:
+ logger.error(f"OFox generation failed before completion: {exc}")
+ _persist_material_sources(task_id, material_sources)
+ raise
+
+ # 单个关键词被远端明确判失败(如触发内容审核)时返回空列表:任务已
+ # 结束、无计费悬念,跳过该片段继续生成后续关键词。
+ for item in video_items:
+ saved_video_path = _save_generated_video_with_retry(
+ item.url, material_directory, "ofox"
+ )
+ if not saved_video_path:
+ # 远端任务已完成并产生费用,本地下载失败时必须把远端任务 ID
+ # 带回任务状态,便于用户去 OFox 控制台找回结果。这里直接抛出
+ # 专用错误,同时阻止后续关键词继续创建新的付费任务。
+ source_info = (
+ item.source_info if isinstance(item.source_info, dict) else {}
+ )
+ remote_task_id = str(source_info.get("asset_id") or "").strip()
+ _persist_material_sources(task_id, material_sources)
+ raise ofox.OFoxDownloadError(
+ "OFox generated a paid video but the result could not be "
+ f"downloaded: id={remote_task_id or 'unknown'}",
+ task_id=remote_task_id,
+ )
+ logger.info(f"video saved: {saved_video_path}")
+ video_paths.append(saved_video_path)
+ try:
+ material_sources.append(_material_source_record(item, saved_video_path))
+ except Exception as source_error:
+ logger.warning(
+ "failed to prepare generated material source record: "
+ f"provider=ofox, "
+ f"error={type(source_error).__name__}, detail={source_error}"
+ )
+ total_duration += min(clip_duration, item.duration)
+ if total_duration >= required_duration:
+ break
+ if total_duration >= required_duration:
+ logger.info(
+ "generated OFox materials cover the required duration; stop "
+ f"submitting paid tasks: generated={total_duration:.1f}s, "
+ f"required={required_duration:.1f}s"
+ )
+ break
+
+ logger.success(f"generated and downloaded {len(video_paths)} OFox videos")
+ _persist_material_sources(task_id, material_sources)
+ return video_paths
+
+
def _download_videos_by_script_order(
task_id: str,
search_terms: List[str],
diff --git a/app/services/ofox.py b/app/services/ofox.py
new file mode 100644
index 00000000..3b0a2397
--- /dev/null
+++ b/app/services/ofox.py
@@ -0,0 +1,441 @@
+import math
+import os
+import time
+from typing import Any, Mapping
+from urllib.parse import quote_plus
+
+import requests
+from loguru import logger
+
+from app.config import config
+from app.models.schema import MaterialInfo, VideoAspect
+
+
+DEFAULT_BASE_URL = "https://api.ofox.ai/v1"
+DEFAULT_MODEL_ID = "bytedance/seedance-2.0-fast"
+DEFAULT_RESOLUTION = "720p"
+# 默认模型 bytedance/seedance-2.0-fast 只接受 4-15 秒(服务端实测校验值)。
+# 其它可选模型区间不同(如 alibaba/wan-2.7 为 2-15 秒),切换模型时应同步
+# 调整配置里的区间;超出区间的请求会被 API 以明确的 400 拒绝,不会计费。
+DEFAULT_MIN_DURATION_SECONDS = 4
+DEFAULT_MAX_DURATION_SECONDS = 15
+DEFAULT_POLL_INTERVAL_SECONDS = 5.0
+DEFAULT_RUN_TIMEOUT_SECONDS = 1800.0
+MAX_POLL_RETRIES = 5
+RETRY_BASE_SECONDS = 1.0
+MAX_ERROR_TEXT_LENGTH = 500
+RETRYABLE_STATUS_CODES = frozenset({429, 500, 502, 503, 504})
+TERMINAL_SUCCESS_STATUSES = frozenset({"completed", "succeeded"})
+TERMINAL_FAILURE_STATUSES = frozenset(
+ {"failed", "error", "cancelled", "canceled", "expired"}
+)
+ACTIVE_STATUSES = frozenset({"queued", "in_progress"})
+
+
+class OFoxError(RuntimeError):
+ """确定性的配置、请求或响应错误。"""
+
+ def __init__(self, message: str, task_id: str = ""):
+ super().__init__(message)
+ # 只要远端任务已经创建,所有错误类型都统一携带任务 ID。上层无需
+ # 根据异常子类分别维护恢复逻辑,WebUI/API 也能稳定展示排障依据。
+ self.task_id = task_id
+
+
+class OFoxUnconfirmedTaskError(OFoxError):
+ """远端可能已创建付费任务,但本机无法确认其最终状态。"""
+
+ def __init__(self, message: str, task_id: str = ""):
+ super().__init__(message, task_id=task_id)
+
+
+class OFoxDownloadError(OFoxError):
+ """远端付费任务已成功,但生成的视频未能下载到本机。"""
+
+ def __init__(self, message: str, task_id: str):
+ super().__init__(message, task_id=task_id)
+
+
+def get_api_key(settings: Mapping[str, Any] | None = None) -> str:
+ """
+ 按明确且唯一的优先级读取 OFox 凭据。
+
+ 配置文件里的专用键优先级最高;唯一支持的运行时环境变量是语义明确的
+ ``OFOX_API_KEY``。
+ """
+ settings = config.app if settings is None else settings
+ configured = str(settings.get("ofox_api_key", "") or "").strip()
+ environment_key = os.getenv("OFOX_API_KEY", "").strip()
+ return configured or environment_key
+
+
+def is_enabled(settings: Mapping[str, Any] | None = None) -> bool:
+ return bool(get_api_key(settings))
+
+
+def _base_url() -> str:
+ return str(
+ config.app.get("ofox_base_url", DEFAULT_BASE_URL) or DEFAULT_BASE_URL
+ ).rstrip("/")
+
+
+def _model_id() -> str:
+ return str(
+ config.app.get("ofox_text_to_video_model", DEFAULT_MODEL_ID) or DEFAULT_MODEL_ID
+ ).strip()
+
+
+def _resolution() -> str:
+ """
+ 读取生成分辨率。
+
+ OFox 按模型在服务端校验分辨率(例如默认 Seedance fast 模型只接受
+ 480p/720p),无效值会得到一个明确的 400 拒绝且不会创建付费任务,
+ 因此这里只做去空白,不维护本地白名单——白名单会随远端模型目录
+ 变化而过期。空值退回默认,避免把空字符串提交到远端。
+ """
+ value = str(config.app.get("ofox_resolution", DEFAULT_RESOLUTION) or "").strip()
+ return value or DEFAULT_RESOLUTION
+
+
+def _config_bool(key: str, default: bool) -> bool:
+ value = config.app.get(key, default)
+ if isinstance(value, str):
+ return value.strip().lower() not in {"0", "false", "no", "off", ""}
+ return bool(value)
+
+
+def _bounded_float(key: str, default: float, minimum: float, maximum: float) -> float:
+ try:
+ value = float(config.app.get(key, default))
+ except (TypeError, ValueError):
+ return default
+ if not math.isfinite(value):
+ return default
+ return min(max(value, minimum), maximum)
+
+
+def _duration_bounds() -> tuple[int, int]:
+ def read(key: str, default: int) -> int:
+ try:
+ value = int(config.app.get(key, default))
+ except (TypeError, ValueError):
+ return default
+ return value if value >= 1 else default
+
+ minimum = read("ofox_min_duration", DEFAULT_MIN_DURATION_SECONDS)
+ maximum = read("ofox_max_duration", DEFAULT_MAX_DURATION_SECONDS)
+ return minimum, max(minimum, maximum)
+
+
+def _tls_verify() -> bool:
+ return _config_bool("tls_verify", True)
+
+
+def _status_code(response: Any) -> int:
+ try:
+ return int(getattr(response, "status_code", 200))
+ except (TypeError, ValueError):
+ return 200
+
+
+def _redact_secret(value: Any, secret: str) -> str:
+ text = str(value or "")
+ if secret:
+ text = text.replace(secret, "***")
+ encoded = quote_plus(secret)
+ if encoded != secret:
+ text = text.replace(encoded, "***")
+ for proxy_url in config.proxy.values():
+ proxy_secret = str(proxy_url or "")
+ if proxy_secret:
+ text = text.replace(proxy_secret, "***")
+ return text[:MAX_ERROR_TEXT_LENGTH]
+
+
+def _response_error(response: Any, api_key: str) -> str:
+ try:
+ payload = response.json()
+ except Exception:
+ return f"HTTP {_status_code(response)}"
+ if not isinstance(payload, dict):
+ return f"HTTP {_status_code(response)}"
+ error = payload.get("error")
+ if isinstance(error, dict):
+ code = error.get("code") or payload.get("code")
+ message = error.get("message") or payload.get("message")
+ else:
+ code = payload.get("code")
+ message = payload.get("message") or error
+ detail = ": ".join(str(item) for item in (code, message) if item not in (None, ""))
+ return _redact_secret(detail or f"HTTP {_status_code(response)}", api_key)
+
+
+def _is_retryable_error(error: Exception) -> bool:
+ if isinstance(
+ error,
+ (
+ requests.exceptions.ConnectionError,
+ requests.exceptions.Timeout,
+ requests.exceptions.ChunkedEncodingError,
+ ),
+ ):
+ return True
+ response = getattr(error, "response", None)
+ return response is not None and _status_code(response) in RETRYABLE_STATUS_CODES
+
+
+def generate_videos(
+ search_term: str,
+ minimum_duration: int,
+ video_aspect: VideoAspect = VideoAspect.portrait,
+) -> list[MaterialInfo]:
+ """提交一个 OFox 文生视频任务,并等待可下载的结果地址。"""
+ api_key = get_api_key()
+ if not api_key:
+ raise OFoxError("OFox video generation requires an OFox API key")
+
+ term = str(search_term or "").strip()
+ if not term:
+ # 空提示词可能来自上游脚本拆分异常。付费生成源不能把它提交到远端,
+ # 否则即使接口接受请求,也只会得到无法使用且已经计费的视频。
+ raise OFoxError("OFox search term must not be empty")
+
+ aspect = VideoAspect(video_aspect)
+ video_width, video_height = aspect.to_resolution()
+ requested_duration = max(int(minimum_duration), 1)
+ minimum, maximum = _duration_bounds()
+ duration = min(max(requested_duration, minimum), maximum)
+ if duration != requested_duration:
+ # 生成比请求更长不会影响成片:剪辑流程仍按片段时长裁剪;生成比请求
+ # 更短只发生在请求超过模型上限时,此时也只能收敛到上限。
+ logger.info(
+ f"ofox clip duration clamped to the configured model range: "
+ f"requested={requested_duration}s, using={duration}s "
+ f"(configured {minimum}-{maximum}s)"
+ )
+ payload = {
+ "model": _model_id(),
+ "prompt": term,
+ "duration": duration,
+ "resolution": _resolution(),
+ "aspect_ratio": aspect.value,
+ }
+ headers = {
+ "Authorization": f"Bearer {api_key}",
+ "Content-Type": "application/json",
+ }
+ videos_url = f"{_base_url()}/videos"
+ logger.info(
+ "generating video with OFox: "
+ f"model={payload['model']}, term={term!r}, duration={duration}s"
+ )
+
+ # 提交接口不做自动重试:超时或 5xx 可能发生在付费任务已经创建之后,
+ # 盲目重试会造成重复扣费。只有拿到明确拒绝响应时才判定为确定性失败。
+ try:
+ response = requests.post(
+ videos_url,
+ json=payload,
+ headers=headers,
+ proxies=config.proxy,
+ verify=_tls_verify(),
+ timeout=(30, 60),
+ )
+ except Exception as exc:
+ raise OFoxUnconfirmedTaskError(
+ "OFox submission returned no response; a paid task may already "
+ "exist remotely: "
+ f"error={type(exc).__name__}, detail={_redact_secret(exc, api_key)}"
+ ) from exc
+
+ status_code = _status_code(response)
+ if status_code >= 500:
+ raise OFoxUnconfirmedTaskError(
+ f"OFox submission failed with HTTP {status_code}; a paid task may "
+ "already exist remotely"
+ )
+ if not 200 <= status_code < 300:
+ # 4xx 是明确拒绝(如 duration/resolution 超出模型支持范围),远端
+ # 没有创建任务,不存在重复计费风险;错误信息里带着服务端给出的
+ # 合法取值范围,直接抛给用户修配置。
+ raise OFoxError(
+ "OFox video generation request rejected: "
+ f"HTTP {status_code}, {_response_error(response, api_key)}"
+ )
+ try:
+ body = response.json()
+ except Exception as exc:
+ raise OFoxUnconfirmedTaskError(
+ "OFox submission returned an unreadable response; a paid task may "
+ f"already exist remotely: error={type(exc).__name__}"
+ ) from exc
+ task_id = str(body.get("id") or "").strip() if isinstance(body, dict) else ""
+ if not task_id:
+ raise OFoxUnconfirmedTaskError(
+ "OFox accepted the submission without returning a task id"
+ )
+ logger.info(f"OFox video task created: id={task_id}")
+
+ task = _wait_for_task(
+ task_id=task_id,
+ videos_url=videos_url,
+ headers=headers,
+ api_key=api_key,
+ )
+ if task is None:
+ return []
+
+ # 产物地址在 unsigned_urls 里,是会过期的临时直链,必须整体保留并
+ # 立即用于下载,不能写入长期的 source_info。
+ unsigned_urls = task.get("unsigned_urls")
+ video_url = ""
+ for candidate in unsigned_urls if isinstance(unsigned_urls, list) else []:
+ if isinstance(candidate, str) and candidate.startswith(("http://", "https://")):
+ video_url = candidate
+ break
+ if not video_url:
+ raise OFoxError(
+ f"OFox task completed without a downloadable video: id={task_id}",
+ task_id=task_id,
+ )
+
+ return [
+ MaterialInfo(
+ provider="ofox",
+ url=video_url,
+ duration=duration,
+ source_info={
+ "provider": "ofox",
+ "search_term": term,
+ "asset_id": task_id,
+ "rendition": {
+ "id": task_id,
+ "width": video_width,
+ "height": video_height,
+ },
+ },
+ )
+ ]
+
+
+def _wait_for_task(
+ *,
+ task_id: str,
+ videos_url: str,
+ headers: dict[str, str],
+ api_key: str,
+) -> dict[str, Any] | None:
+ deadline = time.monotonic() + _bounded_float(
+ "ofox_run_timeout",
+ DEFAULT_RUN_TIMEOUT_SECONDS,
+ 60.0,
+ 7200.0,
+ )
+ poll_interval = _bounded_float(
+ "ofox_poll_interval",
+ DEFAULT_POLL_INTERVAL_SECONDS,
+ 0.5,
+ 60.0,
+ )
+ consecutive_failures = 0
+ while True:
+ remaining = deadline - time.monotonic()
+ if remaining <= 0:
+ raise OFoxUnconfirmedTaskError(
+ "OFox task is still running after the configured local wait "
+ f"timeout: id={task_id}",
+ task_id=task_id,
+ )
+
+ # requests 的 connect/read timeout 分别计时,因此各使用剩余总时间的
+ # 一半。即使连接和读取都走到上限,单轮请求也不会有意超过总截止时间;
+ # 网络库仍可能有极小调度误差,下一处 deadline 检查会阻止再次重试。
+ phase_timeout = max(min(remaining / 2.0, 30.0), 0.001)
+ try:
+ response = requests.get(
+ f"{videos_url}/{task_id}",
+ headers=headers,
+ proxies=config.proxy,
+ verify=_tls_verify(),
+ timeout=(phase_timeout, phase_timeout),
+ )
+ status_code = _status_code(response)
+ if status_code in RETRYABLE_STATUS_CODES:
+ raise requests.exceptions.HTTPError(
+ f"HTTP {status_code}", response=response
+ )
+ if not 200 <= status_code < 300:
+ raise OFoxUnconfirmedTaskError(
+ "OFox task status is unknown: "
+ f"http_status={status_code}, "
+ f"detail={_response_error(response, api_key)}",
+ task_id=task_id,
+ )
+ body = response.json()
+ if not isinstance(body, dict):
+ raise OFoxUnconfirmedTaskError(
+ "OFox task status response is malformed", task_id=task_id
+ )
+ except OFoxUnconfirmedTaskError:
+ raise
+ except Exception as exc:
+ if not _is_retryable_error(exc):
+ raise OFoxUnconfirmedTaskError(
+ "OFox polling failed and the paid task state is unknown: "
+ f"error={type(exc).__name__}, "
+ f"detail={_redact_secret(exc, api_key)}",
+ task_id=task_id,
+ ) from exc
+
+ remaining = deadline - time.monotonic()
+ if remaining <= 0:
+ raise OFoxUnconfirmedTaskError(
+ "OFox task is still running after the configured local wait "
+ f"timeout: id={task_id}",
+ task_id=task_id,
+ ) from exc
+ consecutive_failures += 1
+ if consecutive_failures > MAX_POLL_RETRIES:
+ raise OFoxUnconfirmedTaskError(
+ "OFox polling failed after retries; the paid task may still "
+ f"be running remotely: id={task_id}",
+ task_id=task_id,
+ ) from exc
+ delay = min(RETRY_BASE_SECONDS * consecutive_failures, remaining)
+ logger.warning(
+ "OFox polling hit a transient error; retrying the same task: "
+ f"id={task_id}, attempt={consecutive_failures}/{MAX_POLL_RETRIES}, "
+ f"retry_in={delay:.1f}s"
+ )
+ time.sleep(delay)
+ continue
+
+ consecutive_failures = 0
+ status = str(body.get("status") or "").strip().lower()
+ if status in TERMINAL_SUCCESS_STATUSES:
+ return body
+ if status in TERMINAL_FAILURE_STATUSES:
+ error_detail = body.get("error")
+ logger.error(
+ "OFox task did not produce a video: "
+ f"id={task_id}, status={status}, "
+ f"detail={_redact_secret(error_detail, api_key)}"
+ )
+ # 远端明确失败意味着任务已经结束,可以安全地继续后续片段;由
+ # 调用方决定是否换一个关键词重试。
+ return None
+ if status not in ACTIVE_STATUSES:
+ raise OFoxUnconfirmedTaskError(
+ f"OFox returned an unknown task status: id={task_id}, "
+ f"status={status!r}",
+ task_id=task_id,
+ )
+ remaining = deadline - time.monotonic()
+ if remaining <= 0:
+ raise OFoxUnconfirmedTaskError(
+ "OFox task is still running after the configured local wait "
+ f"timeout: id={task_id}",
+ task_id=task_id,
+ )
+ time.sleep(min(poll_interval, remaining))
diff --git a/app/services/task.py b/app/services/task.py
index a310390d..05a543db 100644
--- a/app/services/task.py
+++ b/app/services/task.py
@@ -20,6 +20,7 @@ from app.services import (
llm,
loomloom,
material,
+ ofox,
sonilo,
subtitle,
task_artifacts,
@@ -739,6 +740,20 @@ def get_video_materials(
details=details,
)
return None
+ except ofox.OFoxError as exc:
+ # 与方舟同一恢复语义:未确认状态和已生成但下载失败都对应一个可在
+ # OFox 控制台恢复的远端任务,统一从异常携带的 task_id 写入失败状态。
+ remote_task_id = str(getattr(exc, "task_id", "") or "").strip()
+ details = (
+ {"ofox_task_id": remote_task_id} if remote_task_id else None
+ )
+ _mark_task_failed(
+ task_id,
+ "materials",
+ str(exc),
+ details=details,
+ )
+ return None
if not downloaded_videos:
_mark_task_failed(
task_id,
@@ -1262,6 +1277,17 @@ def _run_pipeline(
"Volcano Engine Seedance requires an Ark API key",
)
+ if (
+ stop_at in {"materials", "video"}
+ and params.video_source == "ofox"
+ and not ofox.is_enabled()
+ ):
+ return _mark_task_failed(
+ task_id,
+ "preflight",
+ "OFox video generation requires an OFox API key",
+ )
+
if (
stop_at in {"materials", "video"}
and params.video_source == "openai_image"
diff --git a/cli.py b/cli.py
index 9ac602fd..e8b618fa 100644
--- a/cli.py
+++ b/cli.py
@@ -268,6 +268,7 @@ Batch manifests:
"pixabay",
"coverr",
"volcengine_seedance",
+ "ofox",
"openai_image",
"local",
],
@@ -297,6 +298,14 @@ Batch manifests:
"with --video-source volcengine_seedance for materials or video output"
),
)
+ material_group.add_argument(
+ "--confirm-ofox-charge",
+ action="store_true",
+ help=(
+ "confirm that OFox video generation creates paid tasks; required "
+ "with --video-source ofox for materials or video output"
+ ),
+ )
video_group = parser.add_argument_group("video output")
video_group.add_argument(
@@ -595,6 +604,15 @@ Batch manifests:
"--confirm-seedance-charge is required with "
"--video-source volcengine_seedance"
)
+ if (
+ not args.batch_file
+ and args.video_source == "ofox"
+ and stage_requires_materials
+ and not args.confirm_ofox_charge
+ ):
+ parser.error(
+ "--confirm-ofox-charge is required with --video-source ofox"
+ )
if args.bgm_file:
if args.bgm_type in (None, "custom"):
@@ -971,6 +989,7 @@ def _validate_batch_task_params(
stop_at: str,
custom_position_is_explicit: bool,
seedance_charge_confirmed: bool,
+ ofox_charge_confirmed: bool,
) -> None:
if not params.video_subject.strip() and not params.video_script.strip():
raise ValueError("one of video_subject or video_script is required")
@@ -980,11 +999,12 @@ def _validate_batch_task_params(
"pixabay",
"coverr",
"volcengine_seedance",
+ "ofox",
"local",
}:
raise ValueError(
"video_source must be one of: pexels, pixabay, coverr, "
- "volcengine_seedance, local"
+ "volcengine_seedance, ofox, local"
)
for field_name, value in (
("video_aspect", params.video_aspect),
@@ -1018,6 +1038,12 @@ def _validate_batch_task_params(
raise ValueError(
"--confirm-seedance-charge is required for Volcano Engine Seedance"
)
+ if (
+ params.video_source == "ofox"
+ and stop_at in {"materials", "video"}
+ and not ofox_charge_confirmed
+ ):
+ raise ValueError("--confirm-ofox-charge is required for OFox video generation")
if stop_at == "subtitle" and not params.subtitle_enabled:
raise ValueError("stop_at=subtitle cannot be combined with disabled subtitles")
@@ -1118,6 +1144,7 @@ def _build_batch_tasks(args: argparse.Namespace) -> list[VideoParams]:
or "custom_position" in override_fields
),
seedance_charge_confirmed=args.confirm_seedance_charge,
+ ofox_charge_confirmed=args.confirm_ofox_charge,
)
except (TypeError, ValueError) as exc:
raise ValueError(f"invalid batch task {index}: {exc}") from exc
diff --git a/config.example.toml b/config.example.toml
index ef53e49e..de0a5bb8 100644
--- a/config.example.toml
+++ b/config.example.toml
@@ -63,11 +63,14 @@ tls_verify = true
# -----------------------------------------------------------------------------
# Available values: "pexels", "pixabay", "coverr", "wavespeed",
-# "volcengine_seedance", "loomloom", "openai_image", "local".
+# "volcengine_seedance", "ofox", "loomloom", "openai_image", "local".
# "wavespeed" generates brand-new clips with WaveSpeed AI text-to-video models
# instead of searching stock footage; generation is billed per request.
# "loomloom" is a paid Shengsuan Cloud source and requires explicit
# confirmation in the WebUI before each generation request.
+# "ofox" generates brand-new clips through the OFox multi-model video API
+# (Seedance / Wan and other text-to-video models behind one key); generation
+# is billed per submitted clip.
# "openai_image" generates one image per script keyword through any
# OpenAI-compatible /images/generations endpoint (local ComfyUI/SD gateways
# or cloud relays) and renders it into a video clip.
@@ -118,6 +121,28 @@ volcengine_seedance_poll_interval = 5
volcengine_seedance_run_timeout = 1800
volcengine_seedance_watermark = false
+# OFox (https://ofox.ai) exposes multiple text-to-video models (Seedance, Wan,
+# and more) behind one API key using an asynchronous /v1/videos API.
+# Create an API key at https://ofox.ai
+# When empty, the OFOX_API_KEY environment variable is used.
+# Generation is billed per submitted clip.
+ofox_api_key = ""
+ofox_base_url = "https://api.ofox.ai/v1"
+# Any text-to-video model id from the OFox catalog (GET /v1/models) works here.
+ofox_text_to_video_model = "bytedance/seedance-2.0-fast"
+# Supported resolutions depend on the chosen model (the default model accepts
+# "480p" and "720p"); invalid values are rejected by the API with a clear
+# error before any task is created.
+ofox_resolution = "720p"
+# The default model only accepts durations of 4-15 seconds; other models have
+# different ranges (for example alibaba/wan-2.7 accepts 2-15 seconds).
+# Requested clip durations outside the range are clamped automatically before
+# submission (extra footage is trimmed by the editing pipeline).
+ofox_min_duration = 4
+ofox_max_duration = 15
+ofox_poll_interval = 5
+ofox_run_timeout = 1800
+
# OpenAI-compatible text-to-image material source ("openai_image"). Point the
# base URL at any endpoint that speaks the OpenAI /images/generations protocol:
# a local ComfyUI/SD gateway, a self-hosted relay, or a cloud aggregator.
diff --git a/docs/skill/SKILL.md b/docs/skill/SKILL.md
index 9f007716..5452f41e 100644
--- a/docs/skill/SKILL.md
+++ b/docs/skill/SKILL.md
@@ -110,12 +110,17 @@ MPT_LLM_BASE_URL
MPT_LLM_MODEL_NAME
MPT_PEXELS_API_KEY
MPT_VOLCENGINE_ARK_API_KEY
+MPT_OFOX_API_KEY
```
When `SEEDANCE_CHARGE_CONFIRMATION_REQUIRED` is present, explain that every
generated Seedance clip creates a paid Ark task. Only after the user explicitly
confirms, rerun with `--confirm-seedance-charge`; never add this flag silently.
+When `OFOX_CHARGE_CONFIRMATION_REQUIRED` is present, explain that every
+generated OFox clip creates a paid task. Only after the user explicitly
+confirms, rerun with `--confirm-ofox-charge`; never add this flag silently.
+
### Exit code 1: repair or report
Use `MPT_ERROR` and `LOG_FILE` to repair a recoverable problem and retry once. Ask the user only if the repair requires a new API key. If the retry fails, report the failed stage, a short error, and the log path.
diff --git a/docs/skill/mpt_agent.py b/docs/skill/mpt_agent.py
index 928a1e5c..d469fbb9 100644
--- a/docs/skill/mpt_agent.py
+++ b/docs/skill/mpt_agent.py
@@ -30,11 +30,13 @@ SUPPORTED_SOURCES = {
"pixabay",
"coverr",
"volcengine_seedance",
+ "ofox",
"local",
}
VOLCENGINE_ARK_API_KEY_URL = (
"https://console.volcengine.com/ark/region:ark+cn-beijing/apikey"
)
+OFOX_API_KEY_URL = "https://ofox.ai"
PEXELS_API_KEY_URL = "https://www.pexels.com/api/"
PEXELS_VALIDATION_URL = "https://api.pexels.com/v1/collections?per_page=1"
PEXELS_API_KEY_HELP_URL = (
@@ -216,7 +218,10 @@ def apply_environment_config(config_path: Path) -> None:
model_name = os.environ.get("MPT_LLM_MODEL_NAME", "").strip()
pexels_key = os.environ.get("MPT_PEXELS_API_KEY", "").strip()
seedance_key = os.environ.get("MPT_VOLCENGINE_ARK_API_KEY", "").strip()
- if not any((provider, llm_key, base_url, model_name, pexels_key, seedance_key)):
+ ofox_key = os.environ.get("MPT_OFOX_API_KEY", "").strip()
+ if not any(
+ (provider, llm_key, base_url, model_name, pexels_key, seedance_key, ofox_key)
+ ):
return
text = config_path.read_text(encoding="utf-8")
@@ -243,6 +248,9 @@ def apply_environment_config(config_path: Path) -> None:
text, "volcengine_seedance_api_key", seedance_key
)
changes.append("volcengine_seedance_api_key")
+ if ofox_key:
+ text = _replace_config_value(text, "ofox_api_key", ofox_key)
+ changes.append("ofox_api_key")
config_path.write_text(text, encoding="utf-8")
log("updated configuration fields: " + ", ".join(changes))
@@ -334,6 +342,17 @@ def missing_config(config_path: Path, cli_args: list[str]) -> tuple[str, list[st
missing.append("volcengine_seedance_api_key")
if not has_cli_option(cli_args, "--confirm-seedance-charge"):
missing.append("confirm_seedance_charge")
+ elif source == "ofox":
+ # 与运行时 Provider 保持完全一致的凭据优先级:配置键优先,其次是
+ # 语义明确的 OFOX_API_KEY 环境变量。
+ value = (
+ _plain_config_value(text, "ofox_api_key")
+ or os.environ.get("OFOX_API_KEY", "").strip()
+ )
+ if not _has_configured_value(value):
+ missing.append("ofox_api_key")
+ if not has_cli_option(cli_args, "--confirm-ofox-charge"):
+ missing.append("confirm_ofox_charge")
elif source != "local":
value = _plain_config_value(text, f"{source}_api_keys")
if not _has_configured_value(value):
@@ -349,7 +368,7 @@ def report_missing_config(provider: str, missing: list[str]) -> int:
print(f"MISSING={field}")
if any(
field.endswith("_api_key")
- and field != "volcengine_seedance_api_key"
+ and field not in {"volcengine_seedance_api_key", "ofox_api_key"}
for field in missing
):
print("LLM_PROVIDER_OPTIONS_BEGIN")
@@ -373,6 +392,11 @@ def report_missing_config(provider: str, missing: list[str]) -> int:
print("VOLCENGINE_ARK_API_KEY_ENV=MPT_VOLCENGINE_ARK_API_KEY")
if "confirm_seedance_charge" in missing:
print("SEEDANCE_CHARGE_CONFIRMATION_REQUIRED=--confirm-seedance-charge")
+ if "ofox_api_key" in missing:
+ print(f"OFOX_API_KEY_URL={OFOX_API_KEY_URL}")
+ print("OFOX_API_KEY_ENV=MPT_OFOX_API_KEY")
+ if "confirm_ofox_charge" in missing:
+ print("OFOX_CHARGE_CONFIRMATION_REQUIRED=--confirm-ofox-charge")
print("Request only the listed values, set the environment variables, and rerun the same command.")
return NEEDS_INPUT_EXIT_CODE
diff --git a/test/services/test_cli.py b/test/services/test_cli.py
index 66014519..9f3a4987 100644
--- a/test/services/test_cli.py
+++ b/test/services/test_cli.py
@@ -209,6 +209,79 @@ class TestCli(unittest.TestCase):
cli.build_video_params(args).video_source, "volcengine_seedance"
)
+ def test_ofox_video_source_requires_explicit_charge_confirmation(self):
+ with self.assertRaises(SystemExit) as raised:
+ cli.parse_args(
+ ["--video-subject", "test", "--video-source", "ofox"]
+ )
+ self.assertEqual(raised.exception.code, 2)
+
+ args = cli.parse_args(
+ [
+ "--video-subject",
+ "test",
+ "--video-source",
+ "ofox",
+ "--confirm-ofox-charge",
+ ]
+ )
+ self.assertEqual(cli.build_video_params(args).video_source, "ofox")
+
+ def test_ofox_confirmation_is_not_required_before_material_stage(self):
+ args = cli.parse_args(
+ [
+ "--video-subject",
+ "test",
+ "--video-source",
+ "ofox",
+ "--stop-at",
+ "script",
+ ]
+ )
+ self.assertEqual(args.video_source, "ofox")
+
+ def test_batch_ofox_source_uses_global_charge_confirmation(self):
+ with tempfile.TemporaryDirectory() as temp_dir:
+ manifest = Path(temp_dir) / "tasks.json"
+ manifest.write_text(
+ json.dumps(
+ [
+ {
+ "video_subject": "OFox batch task",
+ "video_source": "ofox",
+ }
+ ]
+ ),
+ encoding="utf-8",
+ )
+ with patch("app.services.task.start") as start:
+ rejected = cli.run_cli(
+ ["--batch-file", str(manifest), "--stop-at", "materials"]
+ )
+ self.assertEqual(rejected, 2)
+ start.assert_not_called()
+
+ with (
+ patch(
+ "app.services.task.start",
+ return_value={"state": 1, "materials": ["ok"]},
+ ) as start,
+ patch("app.utils.utils.get_uuid", return_value="task-ofox"),
+ redirect_stdout(io.StringIO()),
+ ):
+ accepted = cli.run_cli(
+ [
+ "--batch-file",
+ str(manifest),
+ "--stop-at",
+ "materials",
+ "--confirm-ofox-charge",
+ ]
+ )
+
+ self.assertEqual(accepted, 0)
+ start.assert_called_once()
+
def test_seedance_confirmation_is_not_required_before_material_stage(self):
args = cli.parse_args(
[
diff --git a/test/services/test_config.py b/test/services/test_config.py
index 5729bc54..6ca3c8ac 100644
--- a/test/services/test_config.py
+++ b/test/services/test_config.py
@@ -42,6 +42,7 @@ class TestConfigPersistence:
"pixabay",
"coverr",
"volcengine_seedance",
+ "ofox",
"loomloom",
"local",
}
@@ -52,6 +53,9 @@ class TestConfigPersistence:
assert app_config["volcengine_seedance_api_key"] == ""
assert app_config["volcengine_seedance_base_url"].startswith("https://")
assert app_config["volcengine_seedance_model"]
+ assert app_config["ofox_api_key"] == ""
+ assert app_config["ofox_base_url"].startswith("https://")
+ assert app_config["ofox_text_to_video_model"]
assert "loomloom_market_listing_id" not in app_config
assert "loomloom_video_market_listing_id" not in app_config
assert app_config["shengsuanyun_api_key"] == ""
diff --git a/test/services/test_mpt_agent_skill.py b/test/services/test_mpt_agent_skill.py
index 58183bc9..af643dd6 100644
--- a/test/services/test_mpt_agent_skill.py
+++ b/test/services/test_mpt_agent_skill.py
@@ -29,6 +29,7 @@ pexels_api_keys = []
pixabay_api_keys = []
coverr_api_keys = []
volcengine_seedance_api_key = ""
+ofox_api_key = ""
oneapi_api_key = ""
oneapi_base_url = ""
oneapi_model_name = ""
@@ -142,6 +143,91 @@ class TestMptAgentSkill(unittest.TestCase):
self.assertEqual(default_missing, ["pexels_api_keys"])
self.assertEqual(pixabay_missing, [])
+ def test_ofox_source_requires_key_and_explicit_charge_confirmation(self):
+ with tempfile.TemporaryDirectory() as temp_dir:
+ config_path = Path(temp_dir) / "config.toml"
+ config_path.write_text(
+ MINIMAL_CONFIG.replace(
+ 'moonshot_api_key = ""', 'moonshot_api_key = "configured"'
+ ),
+ encoding="utf-8",
+ )
+
+ with patch.dict(os.environ, {}, clear=True):
+ _, missing = mpt_agent.missing_config(
+ config_path, ["--video-source", "ofox"]
+ )
+ self.assertEqual(missing, ["ofox_api_key", "confirm_ofox_charge"])
+
+ config_path.write_text(
+ config_path.read_text(encoding="utf-8").replace(
+ 'ofox_api_key = ""', 'ofox_api_key = "configured"'
+ ),
+ encoding="utf-8",
+ )
+ with patch.dict(os.environ, {}, clear=True):
+ _, confirmed_missing = mpt_agent.missing_config(
+ config_path,
+ ["--video-source", "ofox", "--confirm-ofox-charge"],
+ )
+ self.assertEqual(confirmed_missing, [])
+
+ def test_ofox_source_accepts_the_provider_specific_environment_key(self):
+ with tempfile.TemporaryDirectory() as temp_dir:
+ config_path = Path(temp_dir) / "config.toml"
+ config_path.write_text(
+ MINIMAL_CONFIG.replace(
+ 'moonshot_api_key = ""', 'moonshot_api_key = "configured"'
+ ),
+ encoding="utf-8",
+ )
+
+ with patch.dict(
+ os.environ,
+ {"OFOX_API_KEY": "environment-ofox-key"},
+ clear=True,
+ ):
+ _, missing = mpt_agent.missing_config(
+ config_path,
+ ["--video-source", "ofox", "--confirm-ofox-charge"],
+ )
+
+ self.assertEqual(missing, [])
+
+ def test_ofox_environment_key_is_written_without_leaking_to_output(self):
+ with tempfile.TemporaryDirectory() as temp_dir:
+ config_path = Path(temp_dir) / "config.toml"
+ config_path.write_text(MINIMAL_CONFIG, encoding="utf-8")
+ ofox_key = "secret-ofox-key"
+ output = io.StringIO()
+
+ with patch.dict(
+ os.environ,
+ {"MPT_OFOX_API_KEY": ofox_key},
+ clear=True,
+ ), redirect_stdout(output):
+ mpt_agent.apply_environment_config(config_path)
+
+ config = config_path.read_text(encoding="utf-8")
+ self.assertIn(f'ofox_api_key = "{ofox_key}"', config)
+ self.assertNotIn(ofox_key, output.getvalue())
+
+ def test_missing_ofox_inputs_report_signup_and_charge_flag(self):
+ output = io.StringIO()
+ with redirect_stdout(output):
+ code = mpt_agent.report_missing_config(
+ "deepseek",
+ ["ofox_api_key", "confirm_ofox_charge"],
+ )
+
+ text = output.getvalue()
+ self.assertEqual(code, mpt_agent.NEEDS_INPUT_EXIT_CODE)
+ self.assertIn(f"OFOX_API_KEY_URL={mpt_agent.OFOX_API_KEY_URL}", text)
+ self.assertIn(
+ "OFOX_CHARGE_CONFIRMATION_REQUIRED=--confirm-ofox-charge", text
+ )
+ self.assertNotIn("LLM_PROVIDER_OPTIONS_BEGIN", text)
+
def test_seedance_source_requires_key_and_explicit_charge_confirmation(self):
with tempfile.TemporaryDirectory() as temp_dir:
config_path = Path(temp_dir) / "config.toml"
diff --git a/test/services/test_ofox.py b/test/services/test_ofox.py
new file mode 100644
index 00000000..57ab65cf
--- /dev/null
+++ b/test/services/test_ofox.py
@@ -0,0 +1,808 @@
+import os
+import unittest
+from types import SimpleNamespace
+from unittest.mock import patch
+
+import requests
+
+from app.config import config
+from app.models.schema import MaterialInfo, VideoAspect, VideoParams
+from app.services import material, ofox, state as sm, task as task_service
+
+
+class TestOFoxService(unittest.TestCase):
+ def setUp(self):
+ self.original_app = dict(config.app)
+ self.original_proxy = dict(config.proxy)
+ config.app.update(
+ {
+ "ofox_api_key": "ofox-test-key",
+ "ofox_text_to_video_model": ofox.DEFAULT_MODEL_ID,
+ "ofox_base_url": ofox.DEFAULT_BASE_URL,
+ }
+ )
+ config.proxy.clear()
+
+ def tearDown(self):
+ config.app.clear()
+ config.app.update(self.original_app)
+ config.proxy.clear()
+ config.proxy.update(self.original_proxy)
+
+ @staticmethod
+ def _response(payload, status_code=200):
+ return SimpleNamespace(status_code=status_code, json=lambda: payload)
+
+ def test_api_key_prefers_config_then_provider_specific_environment_name(self):
+ with patch.dict(
+ os.environ,
+ {"OFOX_API_KEY": "env-key"},
+ clear=False,
+ ):
+ self.assertEqual(ofox.get_api_key(), "ofox-test-key")
+ config.app["ofox_api_key"] = ""
+ self.assertEqual(ofox.get_api_key(), "env-key")
+ os.environ["OFOX_API_KEY"] = ""
+ self.assertEqual(ofox.get_api_key(), "")
+
+ def test_missing_api_key_fails_before_submission(self):
+ config.app["ofox_api_key"] = ""
+ with (
+ patch.dict(os.environ, {"OFOX_API_KEY": ""}, clear=False),
+ patch.object(ofox.requests, "post") as post,
+ ):
+ with self.assertRaises(ofox.OFoxError):
+ ofox.generate_videos("sunrise", 5)
+ post.assert_not_called()
+
+ def test_empty_search_term_fails_before_paid_submission(self):
+ for invalid in ("", " ", None, 0, False):
+ with self.subTest(invalid=invalid):
+ with patch.object(ofox.requests, "post") as post:
+ with self.assertRaises(ofox.OFoxError) as raised:
+ ofox.generate_videos(invalid, 5)
+
+ self.assertIn("search term must not be empty", str(raised.exception))
+ post.assert_not_called()
+
+ def test_search_term_is_trimmed_consistently_in_request_and_source_record(self):
+ submit = self._response({"id": "vid-trimmed", "status": "queued"}, 202)
+ completed = self._response(
+ {
+ "id": "vid-trimmed",
+ "status": "completed",
+ "unsigned_urls": ["https://cdn.example.com/trimmed.mp4"],
+ }
+ )
+ with (
+ patch.object(ofox.requests, "post", return_value=submit) as post,
+ patch.object(ofox.requests, "get", return_value=completed),
+ ):
+ result = ofox.generate_videos(" smart home ", 5)
+
+ self.assertEqual(post.call_args.kwargs["json"]["prompt"], "smart home")
+ self.assertEqual(result[0].source_info["search_term"], "smart home")
+
+ def test_submit_poll_and_parse_successful_video(self):
+ submit = self._response({"id": "vid-123", "status": "queued"}, 202)
+ polls = [
+ self._response({"id": "vid-123", "status": "in_progress"}),
+ self._response(
+ {
+ "id": "vid-123",
+ "status": "completed",
+ "unsigned_urls": ["https://cdn.example.com/video.mp4?sig=abc"],
+ "usage": {"video_seconds": 5, "video_cost": "0.5"},
+ }
+ ),
+ ]
+
+ with (
+ patch.object(ofox.requests, "post", return_value=submit) as post,
+ patch.object(ofox.requests, "get", side_effect=polls) as get,
+ patch.object(ofox.time, "sleep") as sleep,
+ ):
+ result = ofox.generate_videos(
+ "sunrise over mountains",
+ minimum_duration=5,
+ video_aspect=VideoAspect.portrait,
+ )
+
+ self.assertEqual(len(result), 1)
+ item = result[0]
+ self.assertEqual(item.provider, "ofox")
+ self.assertEqual(item.url, "https://cdn.example.com/video.mp4?sig=abc")
+ self.assertEqual(item.duration, 5)
+ self.assertEqual(item.source_info["asset_id"], "vid-123")
+
+ self.assertEqual(post.call_args.args[0], f"{ofox.DEFAULT_BASE_URL}/videos")
+ self.assertEqual(
+ post.call_args.kwargs["headers"]["Authorization"],
+ "Bearer ofox-test-key",
+ )
+ self.assertEqual(
+ post.call_args.kwargs["json"],
+ {
+ "model": ofox.DEFAULT_MODEL_ID,
+ "prompt": "sunrise over mountains",
+ "duration": 5,
+ "resolution": "720p",
+ "aspect_ratio": "9:16",
+ },
+ )
+ self.assertEqual(get.call_count, 2)
+ self.assertTrue(
+ all("/videos/vid-123" in call.args[0] for call in get.call_args_list)
+ )
+ sleep.assert_called_once_with(ofox.DEFAULT_POLL_INTERVAL_SECONDS)
+
+ def test_resolution_defaults_only_when_blank_and_passes_configured_value(self):
+ config.app.pop("ofox_resolution", None)
+ self.assertEqual(ofox._resolution(), ofox.DEFAULT_RESOLUTION)
+
+ config.app["ofox_resolution"] = " "
+ self.assertEqual(ofox._resolution(), ofox.DEFAULT_RESOLUTION)
+
+ # 分辨率白名单随远端模型目录变化,本地不做校验;配置值原样提交,
+ # 由服务端按模型给出明确的 400 拒绝(不创建付费任务)。
+ config.app["ofox_resolution"] = " 480p "
+ self.assertEqual(ofox._resolution(), "480p")
+
+ def test_server_rejected_resolution_raises_clear_error_without_polling(self):
+ config.app["ofox_resolution"] = "1080p"
+ rejected = self._response(
+ {
+ "error": {
+ "code": "unsupported_parameter",
+ "message": 'resolution "1080p" not supported; allowed: [480p 720p]',
+ }
+ },
+ status_code=400,
+ )
+ with (
+ patch.object(ofox.requests, "post", return_value=rejected),
+ patch.object(ofox.requests, "get") as get,
+ ):
+ with self.assertRaises(ofox.OFoxError) as raised:
+ ofox.generate_videos("city", 5)
+
+ get.assert_not_called()
+ self.assertIn("not supported", str(raised.exception))
+
+ def test_configured_model_base_url_and_duration_bounds_are_applied(self):
+ config.app.update(
+ {
+ "ofox_text_to_video_model": "alibaba/wan-2.7",
+ "ofox_base_url": "https://ofox.example.test/v1/",
+ "ofox_min_duration": 2,
+ "ofox_max_duration": 8,
+ "ofox_resolution": "480p",
+ }
+ )
+ submit = self._response({"id": "vid-bounds", "status": "queued"}, 202)
+ completed = self._response(
+ {
+ "id": "vid-bounds",
+ "status": "completed",
+ "unsigned_urls": ["https://cdn.example.com/bounds.mp4"],
+ }
+ )
+ with (
+ patch.object(ofox.requests, "post", return_value=submit) as post,
+ patch.object(ofox.requests, "get", return_value=completed),
+ ):
+ result = ofox.generate_videos("city", 99, VideoAspect.landscape)
+
+ self.assertEqual(result[0].duration, 8)
+ self.assertEqual(post.call_args.args[0], "https://ofox.example.test/v1/videos")
+ payload = post.call_args.kwargs["json"]
+ self.assertEqual(payload["model"], "alibaba/wan-2.7")
+ self.assertEqual(payload["duration"], 8)
+ self.assertEqual(payload["resolution"], "480p")
+ self.assertEqual(payload["aspect_ratio"], "16:9")
+
+ def test_rejected_submission_raises_clear_error_without_polling(self):
+ rejected = self._response(
+ {
+ "error": {
+ "code": "invalid_request",
+ "message": "rejected ofox-test-key",
+ }
+ },
+ status_code=400,
+ )
+ with (
+ patch.object(ofox.requests, "post", return_value=rejected),
+ patch.object(ofox.requests, "get") as get,
+ ):
+ with self.assertRaises(ofox.OFoxError) as raised:
+ ofox.generate_videos("unsafe", 5)
+
+ get.assert_not_called()
+ self.assertNotIn("ofox-test-key", str(raised.exception))
+
+ def test_submission_network_or_server_error_is_unconfirmed_and_not_retried(self):
+ for side_effect, response in (
+ (requests.exceptions.ConnectionError("offline"), None),
+ (None, self._response({"message": "bad gateway"}, status_code=502)),
+ ):
+ with self.subTest(response=response):
+ with patch.object(
+ ofox.requests,
+ "post",
+ side_effect=side_effect,
+ return_value=response,
+ ) as post:
+ with self.assertRaises(ofox.OFoxUnconfirmedTaskError):
+ ofox.generate_videos("sunrise", 5)
+ self.assertEqual(post.call_count, 1)
+
+ def test_poll_retries_transient_errors_on_same_task(self):
+ submit = self._response({"id": "vid-retry", "status": "queued"}, 202)
+ rate_limited = self._response({}, status_code=429)
+ completed = self._response(
+ {
+ "id": "vid-retry",
+ "status": "completed",
+ "unsigned_urls": ["https://cdn.example.com/retry.mp4"],
+ }
+ )
+ with (
+ patch.object(ofox.requests, "post", return_value=submit) as post,
+ patch.object(
+ ofox.requests,
+ "get",
+ side_effect=[
+ rate_limited,
+ requests.exceptions.ConnectionError("offline"),
+ completed,
+ ],
+ ) as get,
+ patch.object(ofox.time, "sleep") as sleep,
+ ):
+ result = ofox.generate_videos("sunrise", 5)
+
+ self.assertEqual(len(result), 1)
+ self.assertEqual(post.call_count, 1)
+ self.assertEqual(get.call_count, 3)
+ self.assertEqual([call.args[0] for call in sleep.call_args_list], [1.0, 2.0])
+
+ def test_poll_retry_exhaustion_preserves_remote_task_id(self):
+ submit = self._response({"id": "vid-stuck", "status": "queued"}, 202)
+ with (
+ patch.object(ofox.requests, "post", return_value=submit),
+ patch.object(
+ ofox.requests,
+ "get",
+ side_effect=requests.exceptions.ConnectionError("offline"),
+ ) as get,
+ patch.object(ofox.time, "sleep"),
+ ):
+ with self.assertRaises(ofox.OFoxUnconfirmedTaskError) as raised:
+ ofox.generate_videos("sunrise", 5)
+
+ self.assertEqual(raised.exception.task_id, "vid-stuck")
+ self.assertEqual(get.call_count, ofox.MAX_POLL_RETRIES + 1)
+
+ def test_running_task_timeout_preserves_remote_task_id(self):
+ submit = self._response({"id": "vid-running", "status": "queued"}, 202)
+ running = self._response({"id": "vid-running", "status": "in_progress"})
+ clock = iter([0.0, 1.0, ofox.DEFAULT_RUN_TIMEOUT_SECONDS + 1])
+ with (
+ patch.object(ofox.requests, "post", return_value=submit),
+ patch.object(ofox.requests, "get", return_value=running),
+ patch.object(ofox.time, "monotonic", side_effect=lambda: next(clock)),
+ ):
+ with self.assertRaises(ofox.OFoxUnconfirmedTaskError) as raised:
+ ofox.generate_videos("sunrise", 5)
+
+ self.assertEqual(raised.exception.task_id, "vid-running")
+
+ def test_network_retry_stops_when_total_run_deadline_is_reached(self):
+ # 第一次请求前仍有 59 秒;请求异常返回时总截止时间已经过去,不能继续
+ # 执行其余五次网络重试,也不能再进入退避 sleep。
+ config.app["ofox_run_timeout"] = 60
+ clock = iter([0.0, 1.0, 61.0])
+ with (
+ patch.object(
+ ofox.requests,
+ "get",
+ side_effect=requests.exceptions.Timeout("slow"),
+ ) as get,
+ patch.object(ofox.time, "monotonic", side_effect=lambda: next(clock)),
+ patch.object(ofox.time, "sleep") as sleep,
+ ):
+ with self.assertRaises(ofox.OFoxUnconfirmedTaskError) as raised:
+ ofox._wait_for_task(
+ task_id="vid-network-timeout",
+ videos_url="https://ofox.example.test/v1/videos",
+ headers={},
+ api_key="ofox-test-key",
+ )
+
+ self.assertEqual(raised.exception.task_id, "vid-network-timeout")
+ self.assertEqual(get.call_count, 1)
+ self.assertEqual(get.call_args.kwargs["timeout"], (29.5, 29.5))
+ sleep.assert_not_called()
+
+ def test_expired_run_deadline_stops_before_remote_poll(self):
+ config.app["ofox_run_timeout"] = 60
+ clock = iter([0.0, 61.0])
+ with (
+ patch.object(ofox.requests, "get") as get,
+ patch.object(ofox.time, "monotonic", side_effect=lambda: next(clock)),
+ ):
+ with self.assertRaises(ofox.OFoxUnconfirmedTaskError) as raised:
+ ofox._wait_for_task(
+ task_id="vid-expired",
+ videos_url="https://ofox.example.test/v1/videos",
+ headers={},
+ api_key="ofox-test-key",
+ )
+
+ self.assertEqual(raised.exception.task_id, "vid-expired")
+ get.assert_not_called()
+
+ def test_active_poll_sleep_is_capped_by_remaining_deadline(self):
+ config.app.update(
+ {
+ "ofox_run_timeout": 60,
+ "ofox_poll_interval": 5,
+ }
+ )
+ running = self._response({"id": "vid-running", "status": "in_progress"})
+ # 第一次响应完成时只剩 0.25 秒,应只休眠剩余时间;下一轮在发起
+ # 网络请求前发现截止时间已过,避免额外一次远端请求。
+ clock = iter([0.0, 59.0, 59.75, 60.1])
+ with (
+ patch.object(ofox.requests, "get", return_value=running) as get,
+ patch.object(ofox.time, "monotonic", side_effect=lambda: next(clock)),
+ patch.object(ofox.time, "sleep") as sleep,
+ ):
+ with self.assertRaises(ofox.OFoxUnconfirmedTaskError):
+ ofox._wait_for_task(
+ task_id="vid-running",
+ videos_url="https://ofox.example.test/v1/videos",
+ headers={},
+ api_key="ofox-test-key",
+ )
+
+ self.assertEqual(get.call_count, 1)
+ sleep.assert_called_once_with(0.25)
+
+ def test_retry_backoff_is_capped_by_remaining_deadline(self):
+ config.app["ofox_run_timeout"] = 60
+ clock = iter([0.0, 59.0, 59.75, 60.1])
+ with (
+ patch.object(
+ ofox.requests,
+ "get",
+ side_effect=requests.exceptions.Timeout("slow"),
+ ) as get,
+ patch.object(ofox.time, "monotonic", side_effect=lambda: next(clock)),
+ patch.object(ofox.time, "sleep") as sleep,
+ ):
+ with self.assertRaises(ofox.OFoxUnconfirmedTaskError):
+ ofox._wait_for_task(
+ task_id="vid-retry-deadline",
+ videos_url="https://ofox.example.test/v1/videos",
+ headers={},
+ api_key="ofox-test-key",
+ )
+
+ self.assertEqual(get.call_count, 1)
+ self.assertEqual(get.call_args.kwargs["timeout"], (0.5, 0.5))
+ sleep.assert_called_once_with(0.25)
+
+ def test_non_retryable_poll_response_stops_and_redacts_secret(self):
+ unauthorized = self._response(
+ {"error": {"code": "unauthorized", "message": "ofox-test-key"}},
+ status_code=401,
+ )
+ with (
+ patch.object(ofox.requests, "get", return_value=unauthorized) as get,
+ patch.object(ofox.time, "sleep") as sleep,
+ ):
+ with self.assertRaises(ofox.OFoxUnconfirmedTaskError) as raised:
+ ofox._wait_for_task(
+ task_id="vid-unauthorized",
+ videos_url="https://ofox.example.test/v1/videos",
+ headers={},
+ api_key="ofox-test-key",
+ )
+
+ self.assertEqual(raised.exception.task_id, "vid-unauthorized")
+ self.assertNotIn("ofox-test-key", str(raised.exception))
+ self.assertEqual(get.call_count, 1)
+ sleep.assert_not_called()
+
+ def test_malformed_poll_payload_preserves_remote_task_id(self):
+ for payload in ([], None, "in_progress"):
+ with self.subTest(payload=payload):
+ with patch.object(
+ ofox.requests, "get", return_value=self._response(payload)
+ ):
+ with self.assertRaises(ofox.OFoxUnconfirmedTaskError) as raised:
+ ofox._wait_for_task(
+ task_id="vid-malformed",
+ videos_url="https://ofox.example.test/v1/videos",
+ headers={},
+ api_key="ofox-test-key",
+ )
+
+ self.assertEqual(raised.exception.task_id, "vid-malformed")
+
+ def test_all_terminal_failure_statuses_end_the_task_without_billing_doubt(self):
+ # 远端明确失败(如触发内容审核)意味着任务已结束、无计费悬念。返回
+ # None 让上层跳过该关键词继续生成,而不是中止整条视频。
+ for status in ofox.TERMINAL_FAILURE_STATUSES:
+ with self.subTest(status=status):
+ terminal = self._response(
+ {"id": "vid-terminal", "status": status, "error": {}}
+ )
+ with patch.object(ofox.requests, "get", return_value=terminal):
+ result = ofox._wait_for_task(
+ task_id="vid-terminal",
+ videos_url="https://ofox.example.test/v1/videos",
+ headers={},
+ api_key="ofox-test-key",
+ )
+
+ self.assertIsNone(result)
+
+ def test_terminal_failure_returns_empty_result_from_generate(self):
+ submit = self._response({"id": "vid-failed", "status": "queued"}, 202)
+ failed = self._response(
+ {
+ "id": "vid-failed",
+ "status": "failed",
+ "error": {"code": "content_policy", "message": "rejected"},
+ }
+ )
+ with (
+ patch.object(ofox.requests, "post", return_value=submit),
+ patch.object(ofox.requests, "get", return_value=failed),
+ ):
+ result = ofox.generate_videos("sunrise", 5)
+
+ self.assertEqual(result, [])
+
+ def test_unknown_remote_status_is_unconfirmed_and_preserves_task_id(self):
+ unknown = self._response({"id": "vid-unknown", "status": "pausing"})
+ with patch.object(ofox.requests, "get", return_value=unknown):
+ with self.assertRaises(ofox.OFoxUnconfirmedTaskError) as raised:
+ ofox._wait_for_task(
+ task_id="vid-unknown",
+ videos_url="https://ofox.example.test/v1/videos",
+ headers={},
+ api_key="ofox-test-key",
+ )
+
+ self.assertEqual(raised.exception.task_id, "vid-unknown")
+
+ def test_missing_task_id_is_treated_as_unconfirmed(self):
+ for payload in ({}, {"id": " "}, [], None):
+ with self.subTest(payload=payload):
+ with (
+ patch.object(
+ ofox.requests,
+ "post",
+ return_value=self._response(payload, 202),
+ ),
+ patch.object(ofox.requests, "get") as get,
+ ):
+ with self.assertRaises(ofox.OFoxUnconfirmedTaskError):
+ ofox.generate_videos("sunrise", 5)
+ get.assert_not_called()
+
+ def test_completed_task_without_downloadable_url_stops_as_protocol_error(self):
+ for urls in (None, [], ["not-a-url"], [123, {"url": "x"}]):
+ with self.subTest(urls=urls):
+ submit = self._response({"id": "vid-no-url", "status": "queued"}, 202)
+ completed = self._response(
+ {"id": "vid-no-url", "status": "completed", "unsigned_urls": urls}
+ )
+ with (
+ patch.object(ofox.requests, "post", return_value=submit),
+ patch.object(ofox.requests, "get", return_value=completed),
+ ):
+ with self.assertRaises(ofox.OFoxError) as raised:
+ ofox.generate_videos("sunrise", 5)
+
+ self.assertEqual(raised.exception.task_id, "vid-no-url")
+
+ def test_first_valid_unsigned_url_is_selected(self):
+ submit = self._response({"id": "vid-multi", "status": "queued"}, 202)
+ completed = self._response(
+ {
+ "id": "vid-multi",
+ "status": "completed",
+ "unsigned_urls": [
+ None,
+ "ftp://invalid.example.com/a.mp4",
+ "https://cdn.example.com/first-valid.mp4",
+ "https://cdn.example.com/second.mp4",
+ ],
+ }
+ )
+ with (
+ patch.object(ofox.requests, "post", return_value=submit),
+ patch.object(ofox.requests, "get", return_value=completed),
+ ):
+ result = ofox.generate_videos("sunrise", 5)
+
+ self.assertEqual(result[0].url, "https://cdn.example.com/first-valid.mp4")
+
+
+class TestOFoxMaterialIntegration(unittest.TestCase):
+ @staticmethod
+ def _item(term: str, url: str) -> MaterialInfo:
+ return MaterialInfo(
+ provider="ofox",
+ url=url,
+ duration=5,
+ source_info={
+ "provider": "ofox",
+ "search_term": term,
+ "asset_id": f"task-{term}",
+ },
+ )
+
+ def test_on_demand_generation_stops_when_duration_is_covered(self):
+ with (
+ patch.object(
+ ofox,
+ "generate_videos",
+ side_effect=[
+ [self._item("one", "https://cdn.example.com/one.mp4")],
+ [self._item("two", "https://cdn.example.com/two.mp4")],
+ ],
+ ) as generate,
+ patch.object(
+ material,
+ "save_video",
+ side_effect=["/tmp/one.mp4", "/tmp/two.mp4"],
+ ),
+ patch.object(material, "_persist_material_sources") as persist,
+ ):
+ result = material.download_videos(
+ task_id="ofox-materials",
+ search_terms=["one", "two", "three"],
+ source="ofox",
+ audio_duration=10,
+ max_clip_duration=5,
+ )
+
+ self.assertEqual(result, ["/tmp/one.mp4", "/tmp/two.mp4"])
+ self.assertEqual(generate.call_count, 2)
+ self.assertEqual(persist.call_args.args[0], "ofox-materials")
+ self.assertEqual(len(persist.call_args.args[1]), 2)
+
+ def test_failed_keyword_is_skipped_and_generation_continues(self):
+ # 单个关键词被远端明确判失败(generate_videos 返回空列表)时应跳过该
+ # 片段继续下一个关键词,而不是中止整条视频。
+ with (
+ patch.object(
+ ofox,
+ "generate_videos",
+ side_effect=[
+ [],
+ [self._item("two", "https://cdn.example.com/two.mp4")],
+ [self._item("three", "https://cdn.example.com/three.mp4")],
+ ],
+ ) as generate,
+ patch.object(
+ material,
+ "save_video",
+ side_effect=["/tmp/two.mp4", "/tmp/three.mp4"],
+ ),
+ patch.object(material, "_persist_material_sources"),
+ ):
+ result = material.download_videos(
+ task_id="ofox-skip-failed",
+ search_terms=["one", "two", "three"],
+ source="ofox",
+ audio_duration=10,
+ max_clip_duration=5,
+ )
+
+ self.assertEqual(result, ["/tmp/two.mp4", "/tmp/three.mp4"])
+ self.assertEqual(generate.call_count, 3)
+
+ def test_non_positive_audio_duration_avoids_paid_submission(self):
+ for audio_duration in (0, -1, -0.1):
+ with self.subTest(audio_duration=audio_duration):
+ with (
+ patch.object(ofox, "generate_videos") as generate,
+ patch.object(material, "_persist_material_sources") as persist,
+ ):
+ result = material.download_videos(
+ task_id="ofox-no-audio",
+ search_terms=["one", "two"],
+ source="ofox",
+ audio_duration=audio_duration,
+ max_clip_duration=5,
+ )
+
+ self.assertEqual(result, [])
+ generate.assert_not_called()
+ persist.assert_called_once_with("ofox-no-audio", [])
+
+ def test_non_finite_or_invalid_audio_duration_fails_before_paid_submission(self):
+ for audio_duration in (float("nan"), float("inf"), float("-inf"), None, "bad"):
+ with self.subTest(audio_duration=audio_duration):
+ with patch.object(ofox, "generate_videos") as generate:
+ with self.assertRaises(ofox.OFoxError) as raised:
+ material.download_videos(
+ task_id="ofox-invalid-audio",
+ search_terms=["one", "two"],
+ source="ofox",
+ audio_duration=audio_duration,
+ max_clip_duration=5,
+ )
+
+ self.assertIn("audio duration", str(raised.exception))
+ generate.assert_not_called()
+
+ def test_invalid_clip_duration_fails_before_paid_submission(self):
+ for clip_duration in (0, -1, None, "bad", float("nan"), float("inf")):
+ with self.subTest(clip_duration=clip_duration):
+ with patch.object(ofox, "generate_videos") as generate:
+ with self.assertRaises(ofox.OFoxError) as raised:
+ material.download_videos(
+ task_id="ofox-invalid-clip",
+ search_terms=["one", "two"],
+ source="ofox",
+ audio_duration=10,
+ max_clip_duration=clip_duration,
+ )
+
+ self.assertIn("clip duration", str(raised.exception))
+ generate.assert_not_called()
+
+ def test_unconfirmed_task_stops_later_paid_submissions(self):
+ with (
+ patch.object(
+ ofox,
+ "generate_videos",
+ side_effect=ofox.OFoxUnconfirmedTaskError(
+ "unknown", task_id="vid-stuck"
+ ),
+ ) as generate,
+ patch.object(material, "save_video") as save,
+ ):
+ with self.assertRaises(ofox.OFoxUnconfirmedTaskError):
+ material.download_videos(
+ task_id="ofox-unconfirmed",
+ search_terms=["one", "two"],
+ source="ofox",
+ audio_duration=10,
+ max_clip_duration=5,
+ )
+
+ self.assertEqual(generate.call_count, 1)
+ save.assert_not_called()
+
+ def test_download_failure_stops_later_paid_submissions(self):
+ first_item = self._item("one", "https://cdn.example.com/one.mp4")
+ with (
+ patch.object(
+ ofox, "generate_videos", return_value=[first_item]
+ ) as generate,
+ patch.object(material, "_save_generated_video_with_retry", return_value=""),
+ ):
+ with self.assertRaises(ofox.OFoxDownloadError) as raised:
+ material.download_videos(
+ task_id="ofox-download-failed",
+ search_terms=["one", "two"],
+ source="ofox",
+ audio_duration=10,
+ max_clip_duration=5,
+ )
+
+ self.assertEqual(raised.exception.task_id, "task-one")
+ self.assertEqual(generate.call_count, 1)
+
+ def test_later_download_failure_preserves_prior_material_sources(self):
+ first_item = self._item("one", "https://cdn.example.com/one.mp4")
+ second_item = self._item("two", "https://cdn.example.com/two.mp4")
+ with (
+ patch.object(
+ ofox,
+ "generate_videos",
+ side_effect=[[first_item], [second_item]],
+ ) as generate,
+ patch.object(
+ material,
+ "_save_generated_video_with_retry",
+ side_effect=["/tmp/one.mp4", ""],
+ ),
+ patch.object(material, "_persist_material_sources") as persist,
+ ):
+ with self.assertRaises(ofox.OFoxDownloadError) as raised:
+ material.download_videos(
+ task_id="ofox-partial-download",
+ search_terms=["one", "two", "three"],
+ source="ofox",
+ audio_duration=15,
+ max_clip_duration=5,
+ )
+
+ self.assertEqual(raised.exception.task_id, "task-two")
+ self.assertEqual(generate.call_count, 2)
+ persisted_task_id, persisted_sources = persist.call_args.args
+ self.assertEqual(persisted_task_id, "ofox-partial-download")
+ self.assertEqual(len(persisted_sources), 1)
+ self.assertEqual(persisted_sources[0]["asset_id"], "task-one")
+
+ def test_task_preflight_rejects_missing_ofox_key_before_script_generation(self):
+ params = VideoParams(
+ video_subject="OFox preflight",
+ video_source="ofox",
+ )
+ memory_state = sm.MemoryState()
+ with (
+ patch.object(ofox, "is_enabled", return_value=False),
+ patch.object(task_service.sm, "state", memory_state),
+ patch.object(task_service, "generate_script") as generate_script,
+ ):
+ result = task_service.start("ofox-preflight", params, stop_at="materials")
+
+ self.assertEqual(result["failed_stage"], "preflight")
+ self.assertIn("OFox API key", result["error"])
+ generate_script.assert_not_called()
+
+ def test_task_records_ofox_material_error_and_remote_task_id(self):
+ params = VideoParams(
+ video_subject="OFox material failure",
+ video_source="ofox",
+ )
+ memory_state = sm.MemoryState()
+ error = ofox.OFoxUnconfirmedTaskError(
+ "remote state unknown", task_id="vid-recover"
+ )
+ with (
+ patch.object(task_service.sm, "state", memory_state),
+ patch.object(task_service.material, "download_videos", side_effect=error),
+ ):
+ result = task_service.get_video_materials(
+ task_id="ofox-material-error",
+ params=params,
+ video_terms=["scene"],
+ audio_duration=5,
+ )
+
+ self.assertIsNone(result)
+ failed = memory_state.get_task("ofox-material-error")
+ self.assertEqual(failed["failed_stage"], "materials")
+ self.assertEqual(failed["ofox_task_id"], "vid-recover")
+
+ def test_task_records_paid_task_id_when_generated_video_download_fails(self):
+ params = VideoParams(
+ video_subject="OFox download failure",
+ video_source="ofox",
+ )
+ memory_state = sm.MemoryState()
+ error = ofox.OFoxDownloadError(
+ "generated video download failed",
+ task_id="vid-paid-result",
+ )
+ with (
+ patch.object(task_service.sm, "state", memory_state),
+ patch.object(task_service.material, "download_videos", side_effect=error),
+ ):
+ result = task_service.get_video_materials(
+ task_id="ofox-download-error",
+ params=params,
+ video_terms=["scene"],
+ audio_duration=5,
+ )
+
+ self.assertIsNone(result)
+ failed = memory_state.get_task("ofox-download-error")
+ self.assertEqual(failed["failed_stage"], "materials")
+ self.assertEqual(failed["ofox_task_id"], "vid-paid-result")
+
+
+if __name__ == "__main__":
+ unittest.main()
diff --git a/test/services/test_webui_generation_defaults.py b/test/services/test_webui_generation_defaults.py
index 2da6ac15..22955abd 100644
--- a/test/services/test_webui_generation_defaults.py
+++ b/test/services/test_webui_generation_defaults.py
@@ -339,6 +339,29 @@ def test_invalid_saved_generation_settings_fall_back_without_breaking_webui():
assert app.session_state["loomloom_video_scene_count"] == 1
+def test_ofox_source_shows_unchecked_paid_task_confirmation():
+ test_app_config = dict(
+ config.app,
+ video_source="ofox",
+ ofox_api_key="",
+ )
+ test_ui_config = dict(
+ config.ui,
+ language="en",
+ voice_mode="none",
+ )
+
+ with (
+ patch.object(config, "app", test_app_config),
+ patch.object(config, "ui", test_ui_config),
+ patch.object(config, "save_config"),
+ ):
+ app = _new_app()
+
+ assert _widget_by_key(app.selectbox, "video_source_select").value == "ofox"
+ assert _widget_by_key(app.checkbox, "ofox_confirm_charge").value is False
+
+
def test_seedance_source_shows_unchecked_paid_task_confirmation():
test_app_config = dict(
config.app,
diff --git a/test/services/test_webui_ofox.py b/test/services/test_webui_ofox.py
new file mode 100644
index 00000000..b79846f5
--- /dev/null
+++ b/test/services/test_webui_ofox.py
@@ -0,0 +1,58 @@
+from pathlib import Path
+from unittest.mock import patch
+
+from streamlit.testing.v1 import AppTest
+
+from app.config import config
+
+
+ROOT_DIR = Path(__file__).parent.parent.parent
+WEBUI_MAIN = ROOT_DIR / "webui" / "Main.py"
+
+
+def _widget_by_key(elements, key):
+ return next(
+ item
+ for item in elements
+ if str(getattr(item, "key", "")) == key
+ or str(getattr(item, "key", "")).startswith(f"{key}_")
+ )
+
+
+def test_ofox_source_requires_confirmation_then_submits_without_secret_in_params():
+ test_config = dict(
+ config.app,
+ llm_provider="openai",
+ video_source="pexels",
+ ofox_api_key="ofox-secret",
+ )
+ with (
+ patch.object(config, "app", test_config),
+ patch.object(config, "try_save_config", return_value=True),
+ patch("app.services.webui_task.submit_generation") as submit_generation,
+ ):
+ app = AppTest.from_file(str(WEBUI_MAIN), default_timeout=60)
+ app.session_state["ui_language"] = "en"
+ app.run()
+
+ _widget_by_key(app.text_area, "video_subject").set_value("AI office").run()
+ _widget_by_key(app.text_area, "video_script").set_value(
+ "AI helps people work faster."
+ ).run()
+ _widget_by_key(app.text_area, "video_terms").set_value(
+ "office worker, AI assistant"
+ ).run()
+ _widget_by_key(app.selectbox, "video_source_select").select("ofox").run()
+
+ _widget_by_key(app.button, "generate_video_button").click().run()
+ assert submit_generation.call_count == 0
+ assert any("confirm" in str(item.value).lower() for item in app.error)
+
+ _widget_by_key(app.checkbox, "ofox_confirm_charge").check().run()
+ _widget_by_key(app.button, "generate_video_button").click().run()
+
+ assert submit_generation.call_count == 1
+ submitted_params = submit_generation.call_args.kwargs["params"]
+ assert submitted_params.video_source == "ofox"
+ assert "ofox-secret" not in submitted_params.model_dump_json()
+ assert [str(item.value) for item in app.exception] == []
diff --git a/test/services/test_webui_settings_transfer.py b/test/services/test_webui_settings_transfer.py
index 2ac98078..5df80f26 100644
--- a/test/services/test_webui_settings_transfer.py
+++ b/test/services/test_webui_settings_transfer.py
@@ -123,6 +123,7 @@ def _sample_config_sections():
"upload_post_api_key": "api-key-123",
"upload_post_username": "my-username",
"volcengine_seedance_api_key": "ark-seedance-key",
+ "ofox_api_key": "ofox-backup-key",
},
"azure": {"speech_key": "azure-key", "speech_region": "westeurope"},
"elevenlabs": {"api_key": "eleven-key", "model_id": "eleven_v3"},
@@ -225,11 +226,12 @@ def test_key_backup_collects_credentials_and_their_companion_settings():
"upload_post_api_key": "api-key-123",
"upload_post_username": "my-username",
"volcengine_seedance_api_key": "ark-seedance-key",
+ "ofox_api_key": "ofox-backup-key",
},
"azure": {"speech_key": "azure-key", "speech_region": "westeurope"},
"elevenlabs": {"api_key": "eleven-key"},
}
- assert count_backup_keys(backup) == 11
+ assert count_backup_keys(backup) == 12
def test_key_backup_carries_llm_provider_extra_fields_with_the_key():
@@ -284,6 +286,7 @@ def test_key_backup_round_trip_restores_every_saved_key():
assert restored["app"]["upload_post_api_key"] == "api-key-123"
assert restored["app"]["upload_post_username"] == "my-username"
assert restored["app"]["volcengine_seedance_api_key"] == "ark-seedance-key"
+ assert restored["app"]["ofox_api_key"] == "ofox-backup-key"
def test_key_backup_import_ignores_unknown_sections_and_non_key_settings():
@@ -333,6 +336,9 @@ def test_credential_widget_state_keys_match_settings_inputs():
assert credential_widget_state_keys("app", "volcengine_seedance_api_key") == (
"volcengine_seedance_api_key_input",
)
+ assert credential_widget_state_keys("app", "ofox_api_key") == (
+ "ofox_api_key_input",
+ )
assert credential_widget_state_keys("azure", "speech_key") == (
"azure_speech_key_input",
)
diff --git a/webui/Main.py b/webui/Main.py
index ac7b0cfc..e03b3c5e 100644
--- a/webui/Main.py
+++ b/webui/Main.py
@@ -50,6 +50,7 @@ from app.services import (
loomloom,
material,
video,
+ ofox,
volcengine_seedance,
voice,
webui_task,
@@ -106,7 +107,7 @@ LOOMLOOM_MAX_POLL_FAILURES = 5
# config.toml、历史任务和 API 请求中的字段语义,旧用户升级后无需迁移配置。
VIDEO_SOURCE_GROUPS = {
"stock_video": ("pexels", "pixabay", "coverr"),
- "ai_video": ("wavespeed", "volcengine_seedance", "loomloom"),
+ "ai_video": ("wavespeed", "volcengine_seedance", "ofox", "loomloom"),
"ai_image": ("openai_image",),
"local": ("local",),
}
@@ -560,6 +561,7 @@ def _initialize_session_state():
"loomloom_video_confirm_charge": False,
"wavespeed_confirm_charge": False,
"volcengine_seedance_confirm_charge": False,
+ "ofox_confirm_charge": False,
# AI 视频按素材段计费,默认只生成一段,用户确认效果后再主动增加数量。
"loomloom_video_scene_count": _saved_ui_number(
"loomloom_video_scene_count",
@@ -3385,6 +3387,48 @@ def _render_settings_dialog():
seedance_base_url.strip() or volcengine_seedance.DEFAULT_BASE_URL,
)
+ st.markdown("**OFox**")
+ ofox_api_key = st.text_input(
+ tr("OFox API Key"),
+ value=str(config.app.get("ofox_api_key", "") or ""),
+ type="password",
+ key="ofox_api_key_input",
+ )
+ _set_runtime_config("app", "ofox_api_key", ofox_api_key.strip())
+ ofox_model = st.text_input(
+ tr("OFox Text-to-Video Model"),
+ value=str(
+ config.app.get(
+ "ofox_text_to_video_model",
+ ofox.DEFAULT_MODEL_ID,
+ )
+ or ofox.DEFAULT_MODEL_ID
+ ),
+ key="ofox_text_to_video_model_input",
+ )
+ _set_runtime_config(
+ "app", "ofox_text_to_video_model", ofox_model.strip()
+ )
+ configured_ofox_base_url = str(
+ config.app.get("ofox_base_url", ofox.DEFAULT_BASE_URL)
+ or ofox.DEFAULT_BASE_URL
+ ).strip()
+ ofox_base_url = st.text_input(
+ tr("OFox Base URL"),
+ value=(
+ ""
+ if configured_ofox_base_url == ofox.DEFAULT_BASE_URL
+ else configured_ofox_base_url
+ ),
+ placeholder=ofox.DEFAULT_BASE_URL,
+ key="ofox_base_url_input",
+ )
+ _set_runtime_config(
+ "app",
+ "ofox_base_url",
+ ofox_base_url.strip() or ofox.DEFAULT_BASE_URL,
+ )
+
with st.container(border=True):
st.markdown(f"#### {tr('AI Image Generation APIs')}")
st.caption(tr("AI Image Generation APIs Help"))
@@ -4243,6 +4287,7 @@ def _render_video_settings(panel, params):
"coverr": tr("Coverr"),
"wavespeed": tr("WaveSpeed AI Video"),
"volcengine_seedance": tr("Volcano Engine Seedance"),
+ "ofox": tr("OFox AI Video"),
"loomloom": tr("Shengsuan Cloud AI Video"),
"openai_image": tr("OpenAI Compatible Text-to-Image"),
"local": tr("Local file"),
@@ -4270,6 +4315,8 @@ def _render_video_settings(panel, params):
st.caption(tr("WaveSpeed AI Video Help"))
if params.video_source == "volcengine_seedance":
st.caption(tr("Volcano Engine Seedance Help"))
+ if params.video_source == "ofox":
+ st.caption(tr("OFox AI Video Help"))
if params.video_source == "local":
# Streamlit 的文件类型校验对扩展名大小写敏感,这里同时放行大小写两种形式。
local_file_types = sorted(
@@ -4491,6 +4538,8 @@ def _render_video_settings(panel, params):
_render_wavespeed_video_settings(params)
if params.video_source == "volcengine_seedance":
_render_seedance_video_settings(params)
+ if params.video_source == "ofox":
+ _render_ofox_video_settings(params)
return uploaded_files
@@ -4552,6 +4601,30 @@ def _render_seedance_video_settings(params):
)
+def _render_ofox_video_settings(params):
+ """展示预计付费任务数量,并要求用户明确确认 OFox 生成费用。"""
+ clip_duration = max(int(params.video_clip_duration or 1), 1)
+ video_count = max(int(params.video_count or 1), 1)
+ estimated_range = _estimate_voiceover_duration_range(
+ str(params.video_script or ""), params.voice_rate
+ )
+ if estimated_range:
+ min_clips = max(math.ceil(estimated_range[0] * video_count / clip_duration), 1)
+ max_clips = max(
+ math.ceil(estimated_range[1] * video_count / clip_duration), min_clips
+ )
+ st.warning(
+ tr("OFox Billing Notice").format(min=min_clips, max=max_clips)
+ )
+ else:
+ st.warning(tr("OFox Billing Notice Without Script"))
+ st.checkbox(
+ tr("Confirm OFox Charge"),
+ key="ofox_confirm_charge",
+ help=tr("Confirm OFox Charge Help"),
+ )
+
+
def _estimate_voiceover_duration_range(
text: str, voice_rate: float
) -> tuple[float, float] | None:
@@ -6224,6 +6297,7 @@ def _render_generation_controls(
"coverr",
"wavespeed",
"volcengine_seedance",
+ "ofox",
"loomloom",
"openai_image",
"local",
@@ -6283,6 +6357,20 @@ def _render_generation_controls(
st.error(tr("Confirm Volcano Engine Seedance Charge Required"))
st.stop()
+ if params.video_source == "ofox" and not (
+ ofox.is_enabled(config.snapshot_config_with_pending(config.app))
+ ):
+ _remove_active_generation_task(task_id)
+ st.error(tr("Please Enter the OFox API Key"))
+ st.stop()
+
+ if params.video_source == "ofox" and not st.session_state.get(
+ "ofox_confirm_charge", False
+ ):
+ _remove_active_generation_task(task_id)
+ st.error(tr("Confirm OFox Charge Required"))
+ st.stop()
+
if params.video_source == "openai_image" and not material.is_openai_image_enabled(
config.snapshot_config_with_pending(config.app)
):
diff --git a/webui/i18n/de.json b/webui/i18n/de.json
index 4d7ed1f4..e3cf829d 100644
--- a/webui/i18n/de.json
+++ b/webui/i18n/de.json
@@ -375,6 +375,17 @@
"Upload-Post Profile Username": "Upload-Post Profil-Benutzername",
"Upload-Post Profile Username Help": "Geben Sie den unter [Manage Users]({manage_users_url}) angezeigten Benutzernamen ein, nicht die E-Mail-Adresse Ihres Upload-Post-Kontos. Dieser Veröffentlichungsbenutzer muss mit den ausgewählten Plattformen verbunden sein.",
"Upload-Post Setup Guide": "Bevor Sie die automatische Veröffentlichung aktivieren:\n1. Erstellen Sie unter [API Keys]({api_keys_url}) einen API-Schlüssel.\n2. Erstellen oder wählen Sie unter [Manage Users]({manage_users_url}) einen Veröffentlichungsbenutzer und verbinden Sie die Zielplattformen.\n\nGeben Sie unten dessen **Benutzernamen** ein, nicht die E-Mail-Adresse Ihres Upload-Post-Kontos.",
- "YouTube Privacy Status": "YouTube-Datenschutzstatus"
+ "YouTube Privacy Status": "YouTube-Datenschutzstatus",
+ "OFox AI Video": "OFox AI Video",
+ "OFox AI Video Help": "Erzeugen Sie neue Clips über die offizielle OFox API. Jeder eingereichte Clip ist ein kostenpflichtiger OFox-Auftrag.",
+ "OFox API Key": "OFox API-Schlüssel ([API-Schlüssel abrufen](https://ofox.ai))",
+ "OFox Text-to-Video Model": "OFox Text-zu-Video-Modell",
+ "OFox Base URL": "OFox Base-URL",
+ "Please Enter the OFox API Key": "Bitte geben Sie den **OFox API-Schlüssel** ein",
+ "OFox Billing Notice": "Jeder erzeugte Clip wird von OFox berechnet. Auf Grundlage der geschätzten Sprecher- und Clipdauer werden ungefähr {min}–{max} kostenpflichtige Aufträge übermittelt. Die Generierung endet, sobald genügend Material vorhanden ist.",
+ "OFox Billing Notice Without Script": "Jeder erzeugte Clip wird von OFox berechnet. Geben Sie ein Videoskript ein oder erzeugen Sie eines, um die geschätzte Anzahl der Aufträge anzuzeigen.",
+ "Confirm OFox Charge": "Ich verstehe, dass die OFox-Generierung kostenpflichtige OFox-Aufträge erstellt",
+ "Confirm OFox Charge Help": "OFox-Clips werden einzeln übermittelt, bis sie die Dauer der Sprachausgabe abdecken. OFox berechnet jeden eingereichten Generierungsauftrag.",
+ "Confirm OFox Charge Required": "Bitte bestätigen Sie die Kosten der OFox-Generierung, bevor Sie das Video erstellen."
}
}
diff --git a/webui/i18n/en.json b/webui/i18n/en.json
index f9e7cd82..c155f8c5 100644
--- a/webui/i18n/en.json
+++ b/webui/i18n/en.json
@@ -444,6 +444,17 @@
"Upload-Post Profile Username": "Upload-Post Profile Username",
"Upload-Post Profile Username Help": "Enter the Username shown on [Manage Users]({manage_users_url}), not your Upload-Post account email. This publishing user must be connected to the selected platforms.",
"Upload-Post Setup Guide": "Before enabling auto-publish:\n1. Create an API key on [API Keys]({api_keys_url}).\n2. Create or select a publishing user on [Manage Users]({manage_users_url}), then connect the target platforms.\n\nEnter that publishing user's **Username** below, not your Upload-Post account email.",
- "YouTube Privacy Status": "YouTube Privacy Status"
+ "YouTube Privacy Status": "YouTube Privacy Status",
+ "OFox AI Video": "OFox AI Video",
+ "OFox AI Video Help": "Generate brand-new clips through the OFox multi-model video API (Seedance, Wan, and more behind one key). Each submitted clip is a paid task.",
+ "OFox API Key": "OFox API Key ([Get API Key](https://ofox.ai))",
+ "OFox Text-to-Video Model": "OFox Text-to-Video Model",
+ "OFox Base URL": "OFox Base URL",
+ "Please Enter the OFox API Key": "Please enter the **OFox API Key**",
+ "OFox Billing Notice": "Each generated clip is billed by OFox. Based on the estimated voiceover duration and clip duration, roughly {min}-{max} paid task(s) will be submitted. Generation stops as soon as enough footage is collected.",
+ "OFox Billing Notice Without Script": "Each generated clip is billed by OFox. Enter or generate the video script to see an estimated task count.",
+ "Confirm OFox Charge": "I understand OFox video generation creates paid tasks",
+ "Confirm OFox Charge Help": "OFox clips are submitted one at a time and stop once they cover the voiceover duration. OFox bills each submitted generation task.",
+ "Confirm OFox Charge Required": "Please confirm the OFox charge before generating the video."
}
}
diff --git a/webui/i18n/es.json b/webui/i18n/es.json
index 984cfb9a..e1a93ba9 100644
--- a/webui/i18n/es.json
+++ b/webui/i18n/es.json
@@ -375,6 +375,17 @@
"Upload-Post Profile Username": "Nombre de usuario del perfil de Upload-Post",
"Upload-Post Profile Username Help": "Introduce el nombre de usuario que aparece en [Manage Users]({manage_users_url}), no el correo de tu cuenta de Upload-Post. Este usuario de publicación debe estar conectado a las plataformas seleccionadas.",
"Upload-Post Setup Guide": "Antes de activar la publicación automática:\n1. Crea una clave API en [API Keys]({api_keys_url}).\n2. Crea o selecciona un usuario de publicación en [Manage Users]({manage_users_url}) y conecta las plataformas de destino.\n\nIntroduce abajo el **nombre de usuario** de publicación, no el correo de tu cuenta de Upload-Post.",
- "YouTube Privacy Status": "Estado de privacidad de YouTube"
+ "YouTube Privacy Status": "Estado de privacidad de YouTube",
+ "OFox AI Video": "OFox AI Video",
+ "OFox AI Video Help": "Genera clips nuevos mediante la API oficial OFox. Cada clip enviado crea una tarea de pago en OFox.",
+ "OFox API Key": "Clave API de OFox ([Obtener clave API](https://ofox.ai))",
+ "OFox Text-to-Video Model": "Modelo de texto a vídeo de OFox",
+ "OFox Base URL": "URL base de OFox",
+ "Please Enter the OFox API Key": "Introduce la **clave API de OFox**",
+ "OFox Billing Notice": "OFox cobra cada clip generado. Según la duración estimada de la narración y de los clips, se enviarán aproximadamente entre {min} y {max} tareas de pago. La generación se detiene en cuanto haya suficiente material.",
+ "OFox Billing Notice Without Script": "OFox cobra cada clip generado. Introduce o genera el guion del vídeo para ver una estimación del número de tareas.",
+ "Confirm OFox Charge": "Entiendo que la generación con OFox crea tareas de pago en OFox",
+ "Confirm OFox Charge Help": "Los clips de OFox se envían uno a uno y se detienen cuando cubren la duración de la narración. OFox cobra cada tarea de generación enviada.",
+ "Confirm OFox Charge Required": "Confirma el coste de generación de OFox antes de generar el vídeo."
}
}
diff --git a/webui/i18n/fr.json b/webui/i18n/fr.json
index aa2286cd..9ecbdec3 100644
--- a/webui/i18n/fr.json
+++ b/webui/i18n/fr.json
@@ -375,6 +375,17 @@
"Upload-Post Profile Username": "Nom d'utilisateur du profil Upload-Post",
"Upload-Post Profile Username Help": "Saisissez le nom d'utilisateur affiché dans [Manage Users]({manage_users_url}), et non l'adresse e-mail de votre compte Upload-Post. Cet utilisateur de publication doit être connecté aux plateformes sélectionnées.",
"Upload-Post Setup Guide": "Avant d'activer la publication automatique :\n1. Créez une clé API dans [API Keys]({api_keys_url}).\n2. Créez ou sélectionnez un utilisateur de publication dans [Manage Users]({manage_users_url}), puis connectez les plateformes cibles.\n\nSaisissez ci-dessous son **nom d'utilisateur**, et non l'adresse e-mail de votre compte Upload-Post.",
- "YouTube Privacy Status": "Statut de confidentialité YouTube"
+ "YouTube Privacy Status": "Statut de confidentialité YouTube",
+ "OFox AI Video": "OFox AI Video",
+ "OFox AI Video Help": "Générez de nouveaux clips avec l’API officielle OFox. Chaque clip soumis crée une tâche OFox payante.",
+ "OFox API Key": "Clé API OFox ([Obtenir une clé API](https://ofox.ai))",
+ "OFox Text-to-Video Model": "Modèle texte-vidéo OFox",
+ "OFox Base URL": "URL de base OFox",
+ "Please Enter the OFox API Key": "Saisissez la **clé API OFox**",
+ "OFox Billing Notice": "Chaque clip généré est facturé par OFox. D’après la durée estimée de la narration et des clips, environ {min} à {max} tâches payantes seront soumises. La génération s’arrête dès que suffisamment de séquences ont été obtenues.",
+ "OFox Billing Notice Without Script": "Chaque clip généré est facturé par OFox. Saisissez ou générez le script vidéo pour afficher une estimation du nombre de tâches.",
+ "Confirm OFox Charge": "Je comprends que la génération OFox crée des tâches OFox payantes",
+ "Confirm OFox Charge Help": "Les clips OFox sont soumis un par un jusqu’à couvrir la durée de la narration. OFox facture chaque tâche de génération soumise.",
+ "Confirm OFox Charge Required": "Veuillez confirmer les frais de génération OFox avant de créer la vidéo."
}
}
diff --git a/webui/i18n/id.json b/webui/i18n/id.json
index 072e3300..6c11cb02 100644
--- a/webui/i18n/id.json
+++ b/webui/i18n/id.json
@@ -375,6 +375,17 @@
"Upload-Post Profile Username": "Nama Pengguna Profil Upload-Post",
"Upload-Post Profile Username Help": "Masukkan Username yang ditampilkan di [Manage Users]({manage_users_url}), bukan email akun Upload-Post Anda. Pengguna penerbitan ini harus terhubung ke platform yang dipilih.",
"Upload-Post Setup Guide": "Sebelum mengaktifkan publikasi otomatis:\n1. Buat kunci API di [API Keys]({api_keys_url}).\n2. Buat atau pilih pengguna penerbitan di [Manage Users]({manage_users_url}), lalu hubungkan platform tujuan.\n\nMasukkan **Username** pengguna penerbitan tersebut di bawah, bukan email akun Upload-Post Anda.",
- "YouTube Privacy Status": "Status Privasi YouTube"
+ "YouTube Privacy Status": "Status Privasi YouTube",
+ "OFox AI Video": "OFox AI Video",
+ "OFox AI Video Help": "Buat klip baru melalui API resmi OFox. Setiap klip yang dikirim merupakan tugas OFox berbayar.",
+ "OFox API Key": "Kunci API OFox ([Dapatkan Kunci API](https://ofox.ai))",
+ "OFox Text-to-Video Model": "Model teks-ke-video OFox",
+ "OFox Base URL": "URL Dasar OFox",
+ "Please Enter the OFox API Key": "Masukkan **Kunci API OFox**",
+ "OFox Billing Notice": "Setiap klip yang dibuat dikenai biaya oleh OFox. Berdasarkan perkiraan durasi narasi dan klip, sekitar {min}–{max} tugas berbayar akan dikirim. Pembuatan berhenti segera setelah materi mencukupi.",
+ "OFox Billing Notice Without Script": "Setiap klip yang dibuat dikenai biaya oleh OFox. Masukkan atau buat naskah video untuk melihat perkiraan jumlah tugas.",
+ "Confirm OFox Charge": "Saya memahami bahwa pembuatan OFox menghasilkan tugas OFox berbayar",
+ "Confirm OFox Charge Help": "Klip OFox dikirim satu per satu dan berhenti setelah mencakup durasi narasi. OFox mengenakan biaya untuk setiap tugas pembuatan yang dikirim.",
+ "Confirm OFox Charge Required": "Konfirmasikan biaya pembuatan OFox sebelum membuat video."
}
}
diff --git a/webui/i18n/it.json b/webui/i18n/it.json
index 182f20fd..d3ddd0de 100644
--- a/webui/i18n/it.json
+++ b/webui/i18n/it.json
@@ -375,6 +375,17 @@
"Upload-Post Profile Username": "Nome utente del profilo Upload-Post",
"Upload-Post Profile Username Help": "Inserisci il nome utente mostrato in [Manage Users]({manage_users_url}), non l'email del tuo account Upload-Post. Questo utente di pubblicazione deve essere collegato alle piattaforme selezionate.",
"Upload-Post Setup Guide": "Prima di abilitare la pubblicazione automatica:\n1. Crea una chiave API in [API Keys]({api_keys_url}).\n2. Crea o seleziona un utente di pubblicazione in [Manage Users]({manage_users_url}), quindi collega le piattaforme di destinazione.\n\nInserisci qui sotto il suo **nome utente**, non l'email del tuo account Upload-Post.",
- "YouTube Privacy Status": "Stato di privacy di YouTube"
+ "YouTube Privacy Status": "Stato di privacy di YouTube",
+ "OFox AI Video": "OFox AI Video",
+ "OFox AI Video Help": "Genera nuovi clip tramite l’API ufficiale OFox. Ogni clip inviato crea un’attività OFox a pagamento.",
+ "OFox API Key": "Chiave API OFox ([Ottieni chiave API](https://ofox.ai))",
+ "OFox Text-to-Video Model": "Modello testo-video OFox",
+ "OFox Base URL": "URL di base OFox",
+ "Please Enter the OFox API Key": "Inserisci la **chiave API OFox**",
+ "OFox Billing Notice": "Ogni clip generato viene addebitato da OFox. In base alla durata stimata della narrazione e dei clip, verranno inviate circa {min}–{max} attività a pagamento. La generazione si arresta non appena il materiale è sufficiente.",
+ "OFox Billing Notice Without Script": "Ogni clip generato viene addebitato da OFox. Inserisci o genera lo script del video per visualizzare una stima del numero di attività.",
+ "Confirm OFox Charge": "Comprendo che la generazione OFox crea attività OFox a pagamento",
+ "Confirm OFox Charge Help": "I clip OFox vengono inviati uno alla volta e si arrestano quando coprono la durata della narrazione. OFox addebita ogni attività di generazione inviata.",
+ "Confirm OFox Charge Required": "Conferma il costo della generazione OFox prima di creare il video."
}
}
diff --git a/webui/i18n/ko.json b/webui/i18n/ko.json
index 91e83d56..61f38802 100644
--- a/webui/i18n/ko.json
+++ b/webui/i18n/ko.json
@@ -375,6 +375,17 @@
"Upload-Post Profile Username": "Upload-Post 게시 사용자 이름",
"Upload-Post Profile Username Help": "[사용자 관리]({manage_users_url})에 표시된 사용자 이름을 입력하세요. Upload-Post 계정 이메일이 아닙니다. 이 게시 사용자는 선택한 플랫폼에 연결되어 있어야 합니다.",
"Upload-Post Setup Guide": "자동 게시를 활성화하기 전에:\n1. [API 키]({api_keys_url})에서 API 키를 생성하세요.\n2. [사용자 관리]({manage_users_url})에서 게시 사용자를 생성하거나 선택한 다음 대상 플랫폼을 연결하세요.\n\n아래에 Upload-Post 계정 이메일이 아닌 해당 게시 사용자의 **사용자 이름**을 입력하세요.",
- "YouTube Privacy Status": "YouTube 공개 상태"
+ "YouTube Privacy Status": "YouTube 공개 상태",
+ "OFox AI Video": "OFox AI Video",
+ "OFox AI Video Help": "공식 OFox API로 새로운 클립을 생성합니다. 제출하는 각 클립은 유료 OFox 작업입니다.",
+ "OFox API Key": "OFox API 키 ([API 키 발급](https://ofox.ai))",
+ "OFox Text-to-Video Model": "OFox 텍스트-비디오 모델",
+ "OFox Base URL": "OFox 기본 URL",
+ "Please Enter the OFox API Key": "**OFox API 키**를 입력하세요",
+ "OFox Billing Notice": "생성된 각 클립은 OFox에서 과금됩니다. 예상 내레이션 길이와 클립 길이에 따라 약 {min}~{max}개의 유료 작업이 제출됩니다. 필요한 영상이 확보되면 즉시 생성을 중단합니다.",
+ "OFox Billing Notice Without Script": "생성된 각 클립은 OFox에서 과금됩니다. 예상 작업 수를 확인하려면 영상 대본을 입력하거나 생성하세요.",
+ "Confirm OFox Charge": "OFox 생성이 유료 OFox 작업을 만든다는 것을 이해했습니다",
+ "Confirm OFox Charge Help": "OFox 클립은 하나씩 제출되며 내레이션 길이를 채우면 중단됩니다. Ark는 제출된 각 생성 작업에 요금을 부과합니다.",
+ "Confirm OFox Charge Required": "영상을 생성하기 전에 OFox 생성 요금을 확인해 주세요."
}
}
diff --git a/webui/i18n/pt.json b/webui/i18n/pt.json
index 4f27d6ea..dd8e13b7 100644
--- a/webui/i18n/pt.json
+++ b/webui/i18n/pt.json
@@ -375,6 +375,17 @@
"Upload-Post Profile Username": "Nome de Usuário do Perfil Upload-Post",
"Upload-Post Profile Username Help": "Informe o nome de usuário exibido em [Manage Users]({manage_users_url}), não o e-mail da sua conta Upload-Post. Esse usuário de publicação deve estar conectado às plataformas selecionadas.",
"Upload-Post Setup Guide": "Antes de ativar a publicação automática:\n1. Crie uma chave API em [API Keys]({api_keys_url}).\n2. Crie ou selecione um usuário de publicação em [Manage Users]({manage_users_url}) e conecte as plataformas de destino.\n\nInforme abaixo o **nome de usuário** de publicação, não o e-mail da sua conta Upload-Post.",
- "YouTube Privacy Status": "Status de Privacidade do YouTube"
+ "YouTube Privacy Status": "Status de Privacidade do YouTube",
+ "OFox AI Video": "OFox AI Video",
+ "OFox AI Video Help": "Gere novos clipes pela API oficial OFox. Cada clipe enviado cria uma tarefa OFox paga.",
+ "OFox API Key": "Chave de API do OFox ([Obter chave de API](https://ofox.ai))",
+ "OFox Text-to-Video Model": "Modelo de texto para vídeo OFox",
+ "OFox Base URL": "URL base do OFox",
+ "Please Enter the OFox API Key": "Insira a **chave de API do OFox**",
+ "OFox Billing Notice": "Cada clipe gerado é cobrado pelo OFox. Com base na duração estimada da narração e dos clipes, serão enviadas aproximadamente {min}–{max} tarefas pagas. A geração termina assim que houver material suficiente.",
+ "OFox Billing Notice Without Script": "Cada clipe gerado é cobrado pelo OFox. Insira ou gere o roteiro do vídeo para ver uma estimativa do número de tarefas.",
+ "Confirm OFox Charge": "Entendo que a geração OFox cria tarefas OFox pagas",
+ "Confirm OFox Charge Help": "Os clipes OFox são enviados um por vez até cobrirem a duração da narração. O OFox cobra cada tarefa de geração enviada.",
+ "Confirm OFox Charge Required": "Confirme o custo da geração OFox antes de criar o vídeo."
}
}
diff --git a/webui/i18n/ru.json b/webui/i18n/ru.json
index 58a912c6..684665b6 100644
--- a/webui/i18n/ru.json
+++ b/webui/i18n/ru.json
@@ -375,6 +375,17 @@
"Upload-Post Profile Username": "Имя пользователя профиля Upload-Post",
"Upload-Post Profile Username Help": "Введите имя пользователя со страницы [Manage Users]({manage_users_url}), а не адрес электронной почты аккаунта Upload-Post. Этот пользователь публикации должен быть подключён к выбранным платформам.",
"Upload-Post Setup Guide": "Перед включением автопубликации:\n1. Создайте API-ключ на странице [API Keys]({api_keys_url}).\n2. Создайте или выберите пользователя публикации на странице [Manage Users]({manage_users_url}) и подключите целевые платформы.\n\nНиже введите его **имя пользователя**, а не адрес электронной почты аккаунта Upload-Post.",
- "YouTube Privacy Status": "Статус конфиденциальности YouTube"
+ "YouTube Privacy Status": "Статус конфиденциальности YouTube",
+ "OFox AI Video": "OFox AI Video",
+ "OFox AI Video Help": "Создавайте новые клипы через официальный API OFox. Каждый отправленный клип является платной задачей OFox.",
+ "OFox API Key": "API-ключ OFox ([Получить API-ключ](https://ofox.ai))",
+ "OFox Text-to-Video Model": "Модель текст-в-видео OFox",
+ "OFox Base URL": "Базовый URL OFox",
+ "Please Enter the OFox API Key": "Введите **API-ключ OFox**",
+ "OFox Billing Notice": "Каждый созданный клип оплачивается в OFox. С учётом предполагаемой длительности озвучки и клипов будет отправлено примерно {min}–{max} платных задач. Генерация остановится, как только материала будет достаточно.",
+ "OFox Billing Notice Without Script": "Каждый созданный клип оплачивается в OFox. Введите или создайте сценарий видео, чтобы увидеть примерное количество задач.",
+ "Confirm OFox Charge": "Я понимаю, что генерация OFox создаёт платные задачи OFox",
+ "Confirm OFox Charge Help": "Клипы OFox отправляются по одному, пока не покроют длительность озвучки. OFox взимает плату за каждую отправленную задачу генерации.",
+ "Confirm OFox Charge Required": "Подтвердите стоимость генерации OFox перед созданием видео."
}
}
diff --git a/webui/i18n/tr.json b/webui/i18n/tr.json
index b69e8147..b5dbe332 100644
--- a/webui/i18n/tr.json
+++ b/webui/i18n/tr.json
@@ -375,6 +375,17 @@
"Upload-Post Profile Username": "Upload-Post Profil Kullanıcı Adı",
"Upload-Post Profile Username Help": "Upload-Post hesap e-postanız yerine [Manage Users]({manage_users_url}) sayfasında gösterilen kullanıcı adını girin. Bu yayın kullanıcısı seçilen platformlara bağlı olmalıdır.",
"Upload-Post Setup Guide": "Otomatik yayınlamayı etkinleştirmeden önce:\n1. [API Keys]({api_keys_url}) sayfasında bir API anahtarı oluşturun.\n2. [Manage Users]({manage_users_url}) sayfasında bir yayın kullanıcısı oluşturun veya seçin ve hedef platformları bağlayın.\n\nAşağıya Upload-Post hesap e-postanızı değil, bu yayın kullanıcısının **kullanıcı adını** girin.",
- "YouTube Privacy Status": "YouTube Gizlilik Durumu"
+ "YouTube Privacy Status": "YouTube Gizlilik Durumu",
+ "OFox AI Video": "OFox AI Video",
+ "OFox AI Video Help": "Resmî OFox API’siyle yeni klipler oluşturun. Gönderilen her klip ücretli bir OFox görevidir.",
+ "OFox API Key": "OFox API Anahtarı ([API Anahtarı Al](https://ofox.ai))",
+ "OFox Text-to-Video Model": "OFox metinden videoya modeli",
+ "OFox Base URL": "OFox Temel URL’si",
+ "Please Enter the OFox API Key": "Lütfen **OFox API Anahtarını** girin",
+ "OFox Billing Notice": "Oluşturulan her klip OFox tarafından ücretlendirilir. Tahmini seslendirme ve klip süresine göre yaklaşık {min}–{max} ücretli görev gönderilir. Yeterli görüntü elde edildiğinde üretim durur.",
+ "OFox Billing Notice Without Script": "Oluşturulan her klip OFox tarafından ücretlendirilir. Tahmini görev sayısını görmek için video metnini girin veya oluşturun.",
+ "Confirm OFox Charge": "OFox üretiminin ücretli OFox görevleri oluşturduğunu anlıyorum",
+ "Confirm OFox Charge Help": "OFox klipleri tek tek gönderilir ve seslendirme süresini karşıladığında durur. OFox, gönderilen her üretim görevi için ücret alır.",
+ "Confirm OFox Charge Required": "Videoyu oluşturmadan önce OFox üretim ücretini onaylayın."
}
}
diff --git a/webui/i18n/vi.json b/webui/i18n/vi.json
index 43c9c943..ba97fdd2 100644
--- a/webui/i18n/vi.json
+++ b/webui/i18n/vi.json
@@ -375,6 +375,17 @@
"Upload-Post Profile Username": "Tên người dùng hồ sơ Upload-Post",
"Upload-Post Profile Username Help": "Nhập tên người dùng hiển thị tại [Manage Users]({manage_users_url}), không phải email tài khoản Upload-Post. Người dùng xuất bản này phải được kết nối với các nền tảng đã chọn.",
"Upload-Post Setup Guide": "Trước khi bật tự động xuất bản:\n1. Tạo khóa API tại [API Keys]({api_keys_url}).\n2. Tạo hoặc chọn người dùng xuất bản tại [Manage Users]({manage_users_url}), sau đó kết nối các nền tảng đích.\n\nNhập **tên người dùng** xuất bản đó bên dưới, không phải email tài khoản Upload-Post.",
- "YouTube Privacy Status": "Trạng thái quyền riêng tư của YouTube"
+ "YouTube Privacy Status": "Trạng thái quyền riêng tư của YouTube",
+ "OFox AI Video": "OFox AI Video",
+ "OFox AI Video Help": "Tạo clip mới qua API OFox chính thức. Mỗi clip được gửi là một tác vụ OFox có tính phí.",
+ "OFox API Key": "Khóa API OFox ([Lấy khóa API](https://ofox.ai))",
+ "OFox Text-to-Video Model": "Mô hình chuyển văn bản thành video OFox",
+ "OFox Base URL": "URL cơ sở OFox",
+ "Please Enter the OFox API Key": "Vui lòng nhập **Khóa API OFox**",
+ "OFox Billing Notice": "Mỗi clip được tạo đều bị OFox tính phí. Dựa trên thời lượng lồng tiếng và thời lượng clip ước tính, khoảng {min}–{max} tác vụ có tính phí sẽ được gửi. Quá trình tạo dừng ngay khi đã có đủ cảnh quay.",
+ "OFox Billing Notice Without Script": "Mỗi clip được tạo đều bị OFox tính phí. Nhập hoặc tạo kịch bản video để xem số tác vụ ước tính.",
+ "Confirm OFox Charge": "Tôi hiểu rằng việc tạo bằng OFox sẽ tạo các tác vụ OFox có tính phí",
+ "Confirm OFox Charge Help": "Các clip OFox được gửi lần lượt và dừng khi đã đủ thời lượng lồng tiếng. OFox tính phí cho từng tác vụ tạo được gửi.",
+ "Confirm OFox Charge Required": "Vui lòng xác nhận chi phí tạo OFox trước khi tạo video."
}
}
diff --git a/webui/i18n/zh.json b/webui/i18n/zh.json
index 5c81267e..fa212ea7 100644
--- a/webui/i18n/zh.json
+++ b/webui/i18n/zh.json
@@ -444,6 +444,17 @@
"Upload-Post Profile Username": "Upload-Post 发布用户名称",
"Upload-Post Profile Username Help": "填写 [Manage Users]({manage_users_url}) 页面显示的 Username,而不是 Upload-Post 登录邮箱;该发布用户需要已经连接所选平台。",
"Upload-Post Setup Guide": "启用自动发布前,请先:\n1. 在 [API Keys]({api_keys_url}) 页面创建 API Key。\n2. 在 [Manage Users]({manage_users_url}) 页面创建或选择发布用户,并连接目标平台。\n\n下方请填写该发布用户的 **Username**,而不是 Upload-Post 登录邮箱。",
- "YouTube Privacy Status": "YouTube 隐私状态"
+ "YouTube Privacy Status": "YouTube 隐私状态",
+ "OFox AI Video": "OFox AI 视频",
+ "OFox AI Video Help": "通过 OFox 多模型视频 API(一个 Key 可调 Seedance、Wan 等文生视频模型)按脚本关键词生成全新视频素材。每段素材都会创建一个付费任务。",
+ "OFox API Key": "OFox API Key([获取 API Key](https://ofox.ai))",
+ "OFox Text-to-Video Model": "OFox 文生视频模型",
+ "OFox Base URL": "OFox Base URL",
+ "Please Enter the OFox API Key": "请输入 **OFox API Key**",
+ "OFox Billing Notice": "每段生成视频都会由 OFox 计费。根据预计配音时长和片段时长,将提交约 {min}-{max} 个付费任务;凑够所需素材后会立即停止。",
+ "OFox Billing Notice Without Script": "每段生成视频都会由 OFox 计费。填写或生成视频文案后可查看预计任务数量。",
+ "Confirm OFox Charge": "我已了解 OFox 视频生成会创建付费任务",
+ "Confirm OFox Charge Help": "OFox 素材按脚本片段逐段提交,凑够配音所需时长后立即停止。OFox 按实际提交的生成任务计费。",
+ "Confirm OFox Charge Required": "请确认 OFox 生成费用后再生成视频。"
}
}