Skip to content

Use Claude Code, Codex, VSCode, Pi, and OpenCode (and 6 other harnesses) for free (1.3B+ free tokens) from your terminal, app, IDE, or phone, and now from the browser with native browser sessions (multi-harness + multi-model) like OpenClaw (voice supported + ToS friendly)

Python

Add SiliconFlow provider (#1398)

Ali Khokhar19a9d39cf22a
diff --git a/.env.example b/.env.example
index 7432a266..3b729ca8 100644
--- a/.env.example
+++ b/.env.example
@@ -120,6 +120,10 @@ TOGETHER_API_KEY=""
DEEPINFRA_API_KEY=""
+# SiliconFlow (OpenAI-compatible Chat Completions at api.siliconflow.com/v1)
+SILICONFLOW_API_KEY=""
+
+
# Agnes AI (OpenAI-compatible Chat Completions at apihub.agnes-ai.com/v1)
AGNES_API_KEY=""
@@ -158,7 +162,7 @@ OLLAMA_BASE_URL="http://localhost:11434"
# All Claude model requests are mapped to these models, plain model is fallback
# Format: provider_type/model/name
-# Valid providers: "nvidia_nim" | "openai" | "azure_openai" | "open_router" | "gemini" | "vertex" | "deepseek" | "mistral" | "mistral_codestral" | "opencode_zen" | "opencode_go" | "vercel" | "bedrock" | "huggingface" | "cohere" | "github_models" | "wafer" | "kimi" | "kimi_code" | "minimax" | "cerebras" | "groq" | "xai" | "qwencloud" | "together" | "deepinfra" | "agnes" | "zenmux" | "sambanova" | "kilo" | "fireworks" | "novita" | "cloudflare" | "zai" | "ollama_cloud" | "lmstudio" | "llamacpp" | "ollama" | "tokenrouter" | "nararoute"
+# Valid providers: "nvidia_nim" | "openai" | "azure_openai" | "open_router" | "gemini" | "vertex" | "deepseek" | "mistral" | "mistral_codestral" | "opencode_zen" | "opencode_go" | "vercel" | "bedrock" | "huggingface" | "cohere" | "github_models" | "wafer" | "kimi" | "kimi_code" | "minimax" | "cerebras" | "groq" | "xai" | "qwencloud" | "together" | "deepinfra" | "siliconflow" | "agnes" | "zenmux" | "sambanova" | "kilo" | "fireworks" | "novita" | "cloudflare" | "zai" | "ollama_cloud" | "lmstudio" | "llamacpp" | "ollama" | "tokenrouter" | "nararoute"
MODEL_FABLE=
MODEL_OPUS=
MODEL_SONNET=
@@ -203,6 +207,7 @@ FCC_SMOKE_MODEL_GROQ=
FCC_SMOKE_MODEL_QWENCLOUD=
FCC_SMOKE_MODEL_TOGETHER=
FCC_SMOKE_MODEL_DEEPINFRA=
+FCC_SMOKE_MODEL_SILICONFLOW=
FCC_SMOKE_MODEL_AGNES=
FCC_SMOKE_MODEL_ZENMUX=
FCC_SMOKE_MODEL_SAMBANOVA=
@@ -232,6 +237,7 @@ XAI_PROXY=""
QWENCLOUD_PROXY=""
TOGETHER_PROXY=""
DEEPINFRA_PROXY=""
+SILICONFLOW_PROXY=""
AGNES_PROXY=""
ZENMUX_PROXY=""
AZURE_OPENAI_PROXY=""
diff --git a/README.md b/README.md
index 4581de4e..75eed4f1 100644
--- a/README.md
+++ b/README.md
@@ -160,6 +160,7 @@ fcc-codex exec "hello"
| [QwenCloud Token Plan](https://home.qwencloud.com/api-keys) | `QWENCLOUD_API_KEY` | `qwencloud/qwen3.7-plus` |
| [Together AI](https://api.together.ai/settings/api-keys) | `TOGETHER_API_KEY` | `together/zai-org/GLM-5.2` |
| [DeepInfra](https://deepinfra.com/dash/api_keys) | `DEEPINFRA_API_KEY` | `deepinfra/deepseek-ai/DeepSeek-V4-Flash` |
+| [SiliconFlow](https://cloud.siliconflow.com/account/ak) | `SILICONFLOW_API_KEY` | `siliconflow/Qwen/Qwen3-32B` |
| [Agnes AI](https://agnes-ai.com/) | `AGNES_API_KEY` | `agnes/agnes-2.0-flash` |
| [ZenMux](https://zenmux.ai/platform/pay-as-you-go) | `ZENMUX_API_KEY` | `zenmux/deepseek/deepseek-v4-flash-free` |
| [Azure OpenAI](https://learn.microsoft.com/azure/foundry/openai/how-to/chatgpt) | `AZURE_OPENAI_API_KEY` and `AZURE_OPENAI_BASE_URL` | `azure_openai/<deployment-name>` |
diff --git a/pyproject.toml b/pyproject.toml
index b2f27d5e..576a1966 100644
--- a/pyproject.toml
+++ b/pyproject.toml
@@ -4,7 +4,7 @@ build-backend = "hatchling.build"
[project]
name = "free-claude-code"
-version = "4.26.0"
+version = "4.27.0"
description = "Local proxy connecting coding agents to OpenAI-compatible AI providers"
readme = "README.md"
requires-python = ">=3.14.0"
diff --git a/smoke/lib/config.py b/smoke/lib/config.py
index 698a4c39..043ecd44 100644
--- a/smoke/lib/config.py
+++ b/smoke/lib/config.py
@@ -74,6 +74,7 @@ PROVIDER_SMOKE_DEFAULT_MODELS: dict[str, str] = {
"qwencloud": "qwencloud/qwen3.7-plus",
"together": "together/zai-org/GLM-5.2",
"deepinfra": "deepinfra/deepseek-ai/DeepSeek-V4-Flash",
+ "siliconflow": "siliconflow/Qwen/Qwen3-32B",
"sambanova": "sambanova/Meta-Llama-3.3-70B-Instruct",
"kilo": "kilo/kilo-auto/free",
"cerebras": "cerebras/llama3.1-8b",
diff --git a/src/free_claude_code/config/admin/provider_manifest.py b/src/free_claude_code/config/admin/provider_manifest.py
index 547ba3f6..5cbf1310 100644
--- a/src/free_claude_code/config/admin/provider_manifest.py
+++ b/src/free_claude_code/config/admin/provider_manifest.py
@@ -165,6 +165,13 @@ _PROVIDER_FIELD_OVERRIDES: dict[str, dict[str, Any]] = {
"DeepInfra API key for OpenAI-compatible chat and reasoning models."
),
},
+ "SILICONFLOW_API_KEY": {
+ "label": "SiliconFlow API Key",
+ "description": (
+ "SiliconFlow API key for OpenAI-compatible chat, reasoning, and "
+ "vision models."
+ ),
+ },
"SAMBANOVA_API_KEY": {
"label": "SambaNova API Key",
"description": (
diff --git a/src/free_claude_code/config/provider_catalog.py b/src/free_claude_code/config/provider_catalog.py
index b47d49cd..a48bd547 100644
--- a/src/free_claude_code/config/provider_catalog.py
+++ b/src/free_claude_code/config/provider_catalog.py
@@ -60,6 +60,8 @@ QWENCLOUD_DEFAULT_BASE = (
TOGETHER_DEFAULT_BASE = "https://api.together.ai/v1"
# DeepInfra OpenAI-compatible Chat Completions API.
DEEPINFRA_DEFAULT_BASE = "https://api.deepinfra.com/v1/openai"
+# SiliconFlow OpenAI-compatible Chat Completions API.
+SILICONFLOW_DEFAULT_BASE = "https://api.siliconflow.com/v1"
# TokenRouter OpenAI-compatible Chat Completions gateway.
TOKENROUTER_DEFAULT_BASE = "https://api.tokenrouter.com/v1"
# NaraRoute OpenAI-compatible Chat Completions gateway.
@@ -176,6 +178,15 @@ PROVIDER_CATALOG: dict[str, ProviderDescriptor] = {
default_base_url=DEEPINFRA_DEFAULT_BASE,
proxy_attr="deepinfra_proxy",
),
+ "siliconflow": ProviderDescriptor(
+ provider_id="siliconflow",
+ display_name="SiliconFlow",
+ credential_env="SILICONFLOW_API_KEY",
+ credential_url="https://cloud.siliconflow.com/account/ak",
+ credential_attr="siliconflow_api_key",
+ default_base_url=SILICONFLOW_DEFAULT_BASE,
+ proxy_attr="siliconflow_proxy",
+ ),
"agnes": ProviderDescriptor(
provider_id="agnes",
display_name="Agnes AI",
diff --git a/src/free_claude_code/config/settings.py b/src/free_claude_code/config/settings.py
index 08cc014b..13068437 100644
--- a/src/free_claude_code/config/settings.py
+++ b/src/free_claude_code/config/settings.py
@@ -143,6 +143,9 @@ class Settings(BaseSettings):
# ==================== DeepInfra (OpenAI-compatible) ====================
deepinfra_api_key: str = Field(default="", validation_alias="DEEPINFRA_API_KEY")
+ # ==================== SiliconFlow (OpenAI-compatible) ====================
+ siliconflow_api_key: str = Field(default="", validation_alias="SILICONFLOW_API_KEY")
+
# ==================== Agnes AI (OpenAI-compatible) ====================
agnes_api_key: str = Field(default="", validation_alias="AGNES_API_KEY")
@@ -206,6 +209,7 @@ class Settings(BaseSettings):
qwencloud_proxy: str = Field(default="", validation_alias="QWENCLOUD_PROXY")
together_proxy: str = Field(default="", validation_alias="TOGETHER_PROXY")
deepinfra_proxy: str = Field(default="", validation_alias="DEEPINFRA_PROXY")
+ siliconflow_proxy: str = Field(default="", validation_alias="SILICONFLOW_PROXY")
agnes_proxy: str = Field(default="", validation_alias="AGNES_PROXY")
zenmux_proxy: str = Field(default="", validation_alias="ZENMUX_PROXY")
azure_openai_proxy: str = Field(default="", validation_alias="AZURE_OPENAI_PROXY")
diff --git a/src/free_claude_code/providers/openai_chat/extra_body.py b/src/free_claude_code/providers/openai_chat/extra_body.py
index 395ebf30..92cbb369 100644
--- a/src/free_claude_code/providers/openai_chat/extra_body.py
+++ b/src/free_claude_code/providers/openai_chat/extra_body.py
@@ -22,6 +22,8 @@ CANONICAL_OPENAI_CHAT_BODY_KEYS = frozenset(
"reasoning_tokens",
"thinking",
"thinking_budget_tokens",
+ "enable_thinking",
+ "thinking_budget",
}
)
@@ -32,6 +34,8 @@ REASONING_OPENAI_CHAT_BODY_KEYS = frozenset(
"reasoning_tokens",
"thinking",
"thinking_budget_tokens",
+ "enable_thinking",
+ "thinking_budget",
}
)
diff --git a/src/free_claude_code/providers/openai_chat/profiles.py b/src/free_claude_code/providers/openai_chat/profiles.py
index 4fe5cba4..6d28f0a7 100644
--- a/src/free_claude_code/providers/openai_chat/profiles.py
+++ b/src/free_claude_code/providers/openai_chat/profiles.py
@@ -69,6 +69,7 @@ class OpenAIModelListing:
"""Declarative model-list endpoint and response shape."""
path: str | None = None
+ query_params: tuple[tuple[str, str], ...] = ()
collection_field: str | None = "data"
id_field: str = "id"
aliases_field: str | None = None
@@ -237,6 +238,20 @@ OPENAI_CHAT_PROFILES: dict[str, OpenAIChatProfile] = {
non_thinking_tag="non-reasoning",
),
),
+ "siliconflow": OpenAIChatProfile(
+ _policy(
+ "SILICONFLOW",
+ ReasoningReplayMode.REASONING_CONTENT,
+ include_extra_body=True,
+ extra_body_validator=validate_extra_body_does_not_override_reasoning_fields,
+ default_max_tokens=ANTHROPIC_DEFAULT_MAX_OUTPUT_TOKENS,
+ ),
+ NO_REASONING,
+ model_listing=OpenAIModelListing(
+ path="/models",
+ query_params=(("sub_type", "chat"),),
+ ),
+ ),
"agnes": OpenAIChatProfile(
_policy(
"AGNES",
diff --git a/src/free_claude_code/providers/openai_chat/provider.py b/src/free_claude_code/providers/openai_chat/provider.py
index 0ee30c85..9d6b8829 100644
--- a/src/free_claude_code/providers/openai_chat/provider.py
+++ b/src/free_claude_code/providers/openai_chat/provider.py
@@ -169,9 +169,16 @@ class OpenAIChatProvider(BaseProvider):
async def _fetch_models_payload(self) -> Any:
"""Fetch the profile-selected model-list endpoint once."""
- path = self._profile.model_listing.path
+ listing = self._profile.model_listing
+ path = listing.path
if path is None:
return await self._client.models.list()
+ if listing.query_params:
+ return await self._client.get(
+ path,
+ cast_to=object,
+ options={"params": dict(listing.query_params)},
+ )
return await self._client.get(path, cast_to=object)
def _build_request_body(
diff --git a/tests/contracts/test_provider_catalog_order.py b/tests/contracts/test_provider_catalog_order.py
index 70175584..d7dec467 100644
--- a/tests/contracts/test_provider_catalog_order.py
+++ b/tests/contracts/test_provider_catalog_order.py
@@ -14,6 +14,7 @@ _EXPECTED_PROVIDER_ORDER: tuple[str, ...] = (
"qwencloud",
"together",
"deepinfra",
+ "siliconflow",
"agnes",
"zenmux",
"azure_openai",
diff --git a/tests/contracts/test_smoke_config.py b/tests/contracts/test_smoke_config.py
index 0f076890..c004e3fa 100644
--- a/tests/contracts/test_smoke_config.py
+++ b/tests/contracts/test_smoke_config.py
@@ -57,6 +57,7 @@ def _settings(**overrides):
"qwencloud_api_key": "",
"together_api_key": "",
"deepinfra_api_key": "",
+ "siliconflow_api_key": "",
"sambanova_api_key": "",
"cerebras_api_key": "",
"ollama_api_key": "",
@@ -223,6 +224,23 @@ def test_deepinfra_provider_smoke_uses_current_coding_model(monkeypatch) -> None
assert models[0].source == "provider_default"
+def test_siliconflow_provider_smoke_uses_documented_chat_model(monkeypatch) -> None:
+ monkeypatch.delenv("FCC_SMOKE_MODEL_SILICONFLOW", raising=False)
+ config = _smoke_config(
+ settings=_settings(
+ model="ollama/llama3.1",
+ ollama_base_url="",
+ siliconflow_api_key="siliconflow-key",
+ )
+ )
+
+ models = config.provider_smoke_models()
+
+ assert [model.provider for model in models] == ["siliconflow"]
+ assert models[0].full_model == "siliconflow/Qwen/Qwen3-32B"
+ assert models[0].source == "provider_default"
+
+
def test_agnes_provider_smoke_uses_documented_model(monkeypatch) -> None:
monkeypatch.delenv("FCC_SMOKE_MODEL_AGNES", raising=False)
config = _smoke_config(
diff --git a/tests/providers/test_provider_runtime.py b/tests/providers/test_provider_runtime.py
index d21e5f2e..c32b89b7 100644
--- a/tests/providers/test_provider_runtime.py
+++ b/tests/providers/test_provider_runtime.py
@@ -23,6 +23,7 @@ from free_claude_code.config.provider_catalog import (
OLLAMA_CLOUD_DEFAULT_BASE,
PROVIDER_CATALOG,
QWENCLOUD_DEFAULT_BASE,
+ SILICONFLOW_DEFAULT_BASE,
SUPPORTED_PROVIDER_IDS,
TOGETHER_DEFAULT_BASE,
TOKENROUTER_DEFAULT_BASE,
@@ -70,6 +71,7 @@ def _make_settings(**overrides):
mock.qwencloud_api_key = "test_qwencloud_key"
mock.together_api_key = "test_together_key"
mock.deepinfra_api_key = "test_deepinfra_key"
+ mock.siliconflow_api_key = "test_siliconflow_key"
mock.mistral_api_key = "test_mistral_key"
mock.codestral_api_key = "test_codestral_key"
mock.deepseek_api_key = "test_deepseek_key"
@@ -141,6 +143,7 @@ def _make_settings(**overrides):
mock.qwencloud_proxy = ""
mock.together_proxy = ""
mock.deepinfra_proxy = ""
+ mock.siliconflow_proxy = ""
mock.azure_openai_proxy = ""
mock.provider_rate_limit = 40
mock.provider_rate_window = 60
@@ -286,6 +289,26 @@ def test_deepinfra_provider_config_uses_key_base_and_proxy() -> None:
assert isinstance(provider, OpenAIChatProvider)
+def test_siliconflow_provider_config_uses_key_base_and_proxy() -> None:
+ descriptor = PROVIDER_CATALOG["siliconflow"]
+ settings = _make_settings(
+ siliconflow_api_key="siliconflow-token",
+ siliconflow_proxy="http://proxy.test:8080",
+ )
+
+ config = build_provider_config(descriptor, settings)
+ with patch("free_claude_code.providers.openai_chat.provider.AsyncOpenAI"):
+ provider = create_provider("siliconflow", settings)
+
+ assert descriptor.display_name == "SiliconFlow"
+ assert descriptor.credential_env == "SILICONFLOW_API_KEY"
+ assert descriptor.base_url_attr is None
+ assert config.api_key == "siliconflow-token"
+ assert config.base_url == SILICONFLOW_DEFAULT_BASE
+ assert config.proxy == "http://proxy.test:8080"
+ assert isinstance(provider, OpenAIChatProvider)
+
+
def test_agnes_provider_config_uses_key_base_and_proxy() -> None:
descriptor = PROVIDER_CATALOG["agnes"]
settings = _make_settings(
@@ -621,6 +644,7 @@ def test_create_provider_instantiates_each_builtin():
"qwencloud": OpenAIChatProvider,
"together": OpenAIChatProvider,
"deepinfra": OpenAIChatProvider,
+ "siliconflow": OpenAIChatProvider,
"azure_openai": OpenAIChatProvider,
"open_router": OpenRouterProvider,
"mistral": MistralProvider,
diff --git a/tests/providers/test_siliconflow.py b/tests/providers/test_siliconflow.py
new file mode 100644
index 00000000..24bd8ef0
--- /dev/null
+++ b/tests/providers/test_siliconflow.py
@@ -0,0 +1,290 @@
+"""Tests for the SiliconFlow OpenAI-chat profile and model catalog."""
+
+from types import SimpleNamespace
+from typing import Any
+from unittest.mock import AsyncMock, patch
+
+import httpx
+import pytest
+from openai import AsyncOpenAI
+
+from free_claude_code.application.errors import InvalidRequestError
+from free_claude_code.application.model_metadata import ProviderModelInfo
+from free_claude_code.config.constants import ANTHROPIC_DEFAULT_MAX_OUTPUT_TOKENS
+from free_claude_code.config.provider_catalog import SILICONFLOW_DEFAULT_BASE
+from free_claude_code.core.anthropic.models import MessagesRequest
+from free_claude_code.core.reasoning import ReasoningEffort, ReasoningPolicy
+from free_claude_code.providers.base import ProviderConfig
+from free_claude_code.providers.model_listing import ModelListResponseError
+from free_claude_code.providers.openai_chat import OpenAIChatProvider
+from tests.providers.support import (
+ REASONING_OFF,
+ REASONING_ON,
+ immediate_admission,
+ profiled_provider,
+ reasoning_for,
+)
+
+_MODEL = "Qwen/Qwen3-32B"
+_CATALOG_URL = "https://api.siliconflow.com/v1/models?sub_type=chat"
+
+
+def _request(**overrides: Any) -> MessagesRequest:
+ payload: dict[str, Any] = {
+ "model": _MODEL,
+ "messages": [{"role": "user", "content": "Hello"}],
+ }
+ payload.update(overrides)
+ return MessagesRequest.model_validate(payload)
+
+
+def siliconflow_provider() -> OpenAIChatProvider:
+ return profiled_provider(
+ "siliconflow",
+ ProviderConfig(
+ api_key="test-siliconflow-key",
+ base_url=SILICONFLOW_DEFAULT_BASE,
+ rate_limit=10,
+ rate_window=60,
+ ),
+ admission=immediate_admission(provider_name="siliconflow"),
+ )
+
+
+def test_init_uses_documented_endpoint(
+ siliconflow_provider: OpenAIChatProvider,
+) -> None:
+ assert siliconflow_provider._api_key == "test-siliconflow-key"
+ assert siliconflow_provider._base_url == SILICONFLOW_DEFAULT_BASE
+ assert siliconflow_provider._provider_name == "SILICONFLOW"
+
+
+def test_build_request_body_preserves_common_chat_tools_and_images(
+ siliconflow_provider: OpenAIChatProvider,
+) -> None:
+ request = _request(
+ system="Be concise.",
+ messages=[
+ {
+ "role": "user",
+ "content": [
+ {"type": "text", "text": "Inspect this image."},
+ {
+ "type": "image",
+ "source": {
+ "type": "base64",
+ "media_type": "image/png",
+ "data": "aW1hZ2U=",
+ },
+ },
+ ],
+ }
+ ],
+ tools=[
+ {
+ "name": "inspect_image",
+ "description": "Inspect an image",
+ "input_schema": {
+ "type": "object",
+ "properties": {"detail": {"type": "string"}},
+ },
+ }
+ ],
+ )
+
+ body = siliconflow_provider._build_request_body(
+ request,
+ reasoning=ReasoningPolicy.provider_default(),
+ )
+
+ assert body["model"] == _MODEL
+ assert body["max_tokens"] == ANTHROPIC_DEFAULT_MAX_OUTPUT_TOKENS
+ assert body["messages"][0] == {"role": "system", "content": "Be concise."}
+ assert body["messages"][1]["content"] == [
+ {"type": "text", "text": "Inspect this image."},
+ {
+ "type": "image_url",
+ "image_url": {"url": "data:image/png;base64,aW1hZ2U="},
+ },
+ ]
+ assert body["tools"][0]["function"]["name"] == "inspect_image"
+ assert "extra_body" not in body
+
+
+ "reasoning",
+ [
+ ReasoningPolicy.provider_default(),
+ REASONING_OFF,
+ REASONING_ON,
+ ReasoningPolicy.on(budget_tokens=777),
+ *[ReasoningPolicy.on(effort=effort) for effort in ReasoningEffort],
+ ],
+)
+def test_build_request_body_omits_model_specific_thinking_controls(
+ siliconflow_provider: OpenAIChatProvider,
+ reasoning: ReasoningPolicy,
+) -> None:
+ body = siliconflow_provider._build_request_body(_request(), reasoning=reasoning)
+
+ assert "extra_body" not in body
+
+
+def test_build_request_body_preserves_unrelated_extra_body(
+ siliconflow_provider: OpenAIChatProvider,
+) -> None:
+ request = _request(extra_body={"min_p": 0.05})
+
+ body = siliconflow_provider._build_request_body(
+ request,
+ reasoning=ReasoningPolicy.on(effort=ReasoningEffort.HIGH),
+ )
+
+ assert body["extra_body"] == {
+ "min_p": 0.05,
+ }
+
+
[email protected]("field", ["enable_thinking", "thinking_budget"])
+def test_build_request_body_rejects_caller_thinking_override(
+ siliconflow_provider: OpenAIChatProvider,
+ field: str,
+) -> None:
+ request = _request(extra_body={field: "caller-owned"})
+
+ with pytest.raises(InvalidRequestError, match="must not override reasoning"):
+ siliconflow_provider._build_request_body(request, reasoning=REASONING_ON)
+
+
+def test_build_request_body_replays_reasoning_content_verbatim(
+ siliconflow_provider: OpenAIChatProvider,
+) -> None:
+ request = _request(
+ messages=[
+ {"role": "user", "content": "Use a tool."},
+ {
+ "role": "assistant",
+ "content": [
+ {"type": "thinking", "thinking": "Need the tool."},
+ {"type": "text", "text": "Calling it."},
+ ],
+ },
+ {"role": "user", "content": "Continue."},
+ ]
+ )
+
+ body = siliconflow_provider._build_request_body(
+ request,
+ reasoning=reasoning_for(request),
+ )
+
+ assert body["messages"][1] == {
+ "role": "assistant",
+ "content": "Calling it.",
+ "reasoning_content": "Need the tool.",
+ }
+ assert (
+ siliconflow_provider._profile.reasoning_delta(
+ SimpleNamespace(reasoning_content="next thought")
+ )
+ == "next thought"
+ )
+
+
+async def test_model_catalog_uses_chat_query(
+ siliconflow_provider: OpenAIChatProvider,
+) -> None:
+ siliconflow_provider._client.get = AsyncMock(
+ return_value={"data": [{"id": _MODEL}]}
+ )
+
+ model_infos = await siliconflow_provider.list_model_infos()
+
+ assert model_infos == frozenset({ProviderModelInfo(_MODEL)})
+ siliconflow_provider._client.get.assert_awaited_once_with(
+ "/models",
+ cast_to=object,
+ options={"params": {"sub_type": "chat"}},
+ )
+
+
+async def test_model_catalog_uses_documented_endpoint_query_and_auth() -> None:
+ requests: list[httpx.Request] = []
+
+ def handler(request: httpx.Request) -> httpx.Response:
+ requests.append(request)
+ return httpx.Response(200, json={"data": [{"id": _MODEL}]})
+
+ http_client = httpx.AsyncClient(transport=httpx.MockTransport(handler))
+
+ def build_client(*args: Any, **kwargs: Any) -> AsyncOpenAI:
+ kwargs["http_client"] = http_client
+ return AsyncOpenAI(*args, **kwargs)
+
+ with patch(
+ "free_claude_code.providers.openai_chat.provider.AsyncOpenAI",
+ side_effect=build_client,
+ ):
+ provider = profiled_provider(
+ "siliconflow",
+ ProviderConfig(
+ api_key="wire-siliconflow-key",
+ base_url=SILICONFLOW_DEFAULT_BASE,
+ rate_limit=10,
+ rate_window=60,
+ ),
+ admission=immediate_admission(provider_name="siliconflow"),
+ )
+ try:
+ model_infos = await provider.list_model_infos()
+ finally:
+ await provider.cleanup()
+
+ assert model_infos == frozenset({ProviderModelInfo(_MODEL)})
+ assert len(requests) == 1
+ assert str(requests[0].url) == _CATALOG_URL
+ assert requests[0].headers["authorization"] == "Bearer wire-siliconflow-key"
+
+
+ ("payload", "message"),
+ [
+ ({"models": []}, "expected top-level data array"),
+ ({"data": [{}]}, "include id"),
+ ({"data": []}, "did not include any model ids"),
+ ],
+)
+async def test_rejects_malformed_or_empty_catalog_atomically(
+ siliconflow_provider: OpenAIChatProvider,
+ payload: object,
+ message: str,
+) -> None:
+ siliconflow_provider._client.get = AsyncMock(return_value=payload)
+
+ with pytest.raises(ModelListResponseError, match=message):
+ await siliconflow_provider.list_model_infos()
+
+
+async def test_model_catalog_uses_shared_retry_admission(
+ siliconflow_provider: OpenAIChatProvider,
+) -> None:
+ request = httpx.Request("GET", _CATALOG_URL)
+ response = httpx.Response(429, request=request)
+ rate_limit_error = httpx.HTTPStatusError(
+ "rate limited",
+ request=request,
+ response=response,
+ )
+ siliconflow_provider._client.get = AsyncMock(
+ side_effect=[rate_limit_error, {"data": [{"id": _MODEL}]}]
+ )
+
+ model_infos = await siliconflow_provider.list_model_infos()
+
+ assert model_infos == frozenset({ProviderModelInfo(_MODEL)})
+ assert siliconflow_provider._client.get.await_count == 2
diff --git a/uv.lock b/uv.lock
index 1fdd2931..9be040bf 100644
--- a/uv.lock
+++ b/uv.lock
@@ -620,7 +620,7 @@ wheels = [
[[package]]
name = "free-claude-code"
-version = "4.26.0"
+version = "4.27.0"
source = { editable = "." }
dependencies = [
{ name = "aiohttp" },