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 Agnes AI provider (#1176) (#1387)

Tanishq-1cddbaadf5e76
diff --git a/.env.example b/.env.example
index fd308270..eed486a5 100644
--- a/.env.example
+++ b/.env.example
@@ -121,6 +121,10 @@ TOGETHER_API_KEY=""
DEEPINFRA_API_KEY=""
+# Agnes AI (OpenAI-compatible Chat Completions at apihub.agnes-ai.com/v1)
+AGNES_API_KEY=""
+
+
# SambaNova Cloud (OpenAI-compatible Chat Completions at api.sambanova.ai/v1)
SAMBANOVA_API_KEY=""
@@ -151,7 +155,7 @@ 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" | "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" | "agnes" | "sambanova" | "kilo" | "fireworks" | "novita" | "cloudflare" | "zai" | "ollama_cloud" | "lmstudio" | "llamacpp" | "ollama" | "tokenrouter" | "nararoute"
MODEL_FABLE=
MODEL_OPUS=
MODEL_SONNET=
@@ -196,6 +200,7 @@ FCC_SMOKE_MODEL_GROQ=
FCC_SMOKE_MODEL_QWENCLOUD=
FCC_SMOKE_MODEL_TOGETHER=
FCC_SMOKE_MODEL_DEEPINFRA=
+FCC_SMOKE_MODEL_AGNES=
FCC_SMOKE_MODEL_SAMBANOVA=
FCC_SMOKE_MODEL_KILO=
FCC_SMOKE_MODEL_CEREBRAS=
@@ -223,6 +228,7 @@ XAI_PROXY=""
QWENCLOUD_PROXY=""
TOGETHER_PROXY=""
DEEPINFRA_PROXY=""
+AGNES_PROXY=""
AZURE_OPENAI_PROXY=""
NVIDIA_NIM_PROXY=""
OPENROUTER_PROXY=""
diff --git a/README.md b/README.md
index 6ee70f98..ae820ea1 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` |
+| [Agnes AI](https://agnes-ai.com/) | `AGNES_API_KEY` | `agnes/agnes-2.0-flash` |
| [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>` |
| [Google AI Studio (Gemini)](https://aistudio.google.com/apikey) | `GEMINI_API_KEY` | `gemini/models/gemini-3.1-flash-lite` |
| [Google Vertex AI](https://cloud.google.com/vertex-ai/generative-ai/docs/start/openai) | `VERTEX_PROJECT_ID` + ADC | `vertex/google/gemini-3.5-flash` |
diff --git a/pyproject.toml b/pyproject.toml
index 34866e1f..01983cbb 100644
--- a/pyproject.toml
+++ b/pyproject.toml
@@ -4,7 +4,7 @@ build-backend = "hatchling.build"
[project]
name = "free-claude-code"
-version = "4.24.0"
+version = "4.25.0"
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 0e162f25..0783b80e 100644
--- a/smoke/lib/config.py
+++ b/smoke/lib/config.py
@@ -81,6 +81,7 @@ PROVIDER_SMOKE_DEFAULT_MODELS: dict[str, str] = {
"cloudflare": "cloudflare/@cf/moonshotai/kimi-k2.6",
"tokenrouter": "tokenrouter/moonshotai/kimi-k3-free",
"nararoute": "nararoute/kimi-k3-free",
+ "agnes": "agnes/agnes-2.0-flash",
}
MISTRAL_REASONING_SMOKE_DEFAULT_MODEL = "mistral/mistral-medium-3-5"
diff --git a/src/free_claude_code/config/admin/provider_manifest.py b/src/free_claude_code/config/admin/provider_manifest.py
index d296e4e9..b9d508a4 100644
--- a/src/free_claude_code/config/admin/provider_manifest.py
+++ b/src/free_claude_code/config/admin/provider_manifest.py
@@ -210,6 +210,12 @@ _PROVIDER_FIELD_OVERRIDES: dict[str, dict[str, Any]] = {
"Defaults to https://router.bynara.id/v1."
),
},
+ "AGNES_API_KEY": {
+ "label": "Agnes AI API Key",
+ "description": (
+ "Agnes AI OpenAI-compatible API key for apihub.agnes-ai.com/v1."
+ ),
+ },
}
diff --git a/src/free_claude_code/config/provider_catalog.py b/src/free_claude_code/config/provider_catalog.py
index 6158a250..1852c8fc 100644
--- a/src/free_claude_code/config/provider_catalog.py
+++ b/src/free_claude_code/config/provider_catalog.py
@@ -64,6 +64,8 @@ DEEPINFRA_DEFAULT_BASE = "https://api.deepinfra.com/v1/openai"
TOKENROUTER_DEFAULT_BASE = "https://api.tokenrouter.com/v1"
# NaraRoute OpenAI-compatible Chat Completions gateway.
NARAROUTE_DEFAULT_BASE = "https://router.bynara.id/v1"
+# Agnes AI OpenAI-compatible Chat Completions API.
+AGNES_DEFAULT_BASE = "https://apihub.agnes-ai.com/v1"
class ProviderAuthKind(StrEnum):
@@ -172,6 +174,15 @@ PROVIDER_CATALOG: dict[str, ProviderDescriptor] = {
default_base_url=DEEPINFRA_DEFAULT_BASE,
proxy_attr="deepinfra_proxy",
),
+ "agnes": ProviderDescriptor(
+ provider_id="agnes",
+ display_name="Agnes AI",
+ credential_env="AGNES_API_KEY",
+ credential_url="https://agnes-ai.com/",
+ credential_attr="agnes_api_key",
+ default_base_url=AGNES_DEFAULT_BASE,
+ proxy_attr="agnes_proxy",
+ ),
"azure_openai": ProviderDescriptor(
provider_id="azure_openai",
display_name="Azure OpenAI",
diff --git a/src/free_claude_code/config/settings.py b/src/free_claude_code/config/settings.py
index 31be1520..ad0bebad 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")
+ # ==================== Agnes AI (OpenAI-compatible) ====================
+ agnes_api_key: str = Field(default="", validation_alias="AGNES_API_KEY")
+
# ==================== Cerebras Inference (OpenAI-compatible) ====================
cerebras_api_key: str = Field(default="", validation_alias="CEREBRAS_API_KEY")
@@ -200,6 +203,7 @@ 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")
+ agnes_proxy: str = Field(default="", validation_alias="AGNES_PROXY")
azure_openai_proxy: str = Field(default="", validation_alias="AZURE_OPENAI_PROXY")
nvidia_nim_proxy: str = Field(default="", validation_alias="NVIDIA_NIM_PROXY")
open_router_proxy: str = Field(default="", validation_alias="OPENROUTER_PROXY")
diff --git a/src/free_claude_code/providers/openai_chat/profiles.py b/src/free_claude_code/providers/openai_chat/profiles.py
index da8df8de..92863399 100644
--- a/src/free_claude_code/providers/openai_chat/profiles.py
+++ b/src/free_claude_code/providers/openai_chat/profiles.py
@@ -20,6 +20,7 @@ from .reasoning import (
LLAMACPP_REASONING,
NO_REASONING,
SPLIT_REASONING_OUTPUT,
+ ChatTemplateReasoning,
NamedEffortReasoning,
ReasoningEncoder,
ReasoningObject,
@@ -214,6 +215,16 @@ OPENAI_CHAT_PROFILES: dict[str, OpenAIChatProfile] = {
non_thinking_tag="non-reasoning",
),
),
+ "agnes": OpenAIChatProfile(
+ _policy(
+ "AGNES",
+ ReasoningReplayMode.THINK_TAGS,
+ include_extra_body=True,
+ extra_body_validator=validate_extra_body_does_not_override_reasoning_fields,
+ default_max_tokens=ANTHROPIC_DEFAULT_MAX_OUTPUT_TOKENS,
+ ),
+ ChatTemplateReasoning(field="enable_thinking"),
+ ),
"azure_openai": OpenAIChatProfile(
_policy(
"AZURE_OPENAI",
diff --git a/tests/contracts/test_provider_catalog_order.py b/tests/contracts/test_provider_catalog_order.py
index 7e6cfebb..6edda527 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",
+ "agnes",
"azure_openai",
"gemini",
"vertex",
diff --git a/tests/contracts/test_smoke_config.py b/tests/contracts/test_smoke_config.py
index db4c8efb..d917ff75 100644
--- a/tests/contracts/test_smoke_config.py
+++ b/tests/contracts/test_smoke_config.py
@@ -223,6 +223,23 @@ def test_deepinfra_provider_smoke_uses_current_coding_model(monkeypatch) -> None
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(
+ settings=_settings(
+ model="ollama/llama3.1",
+ ollama_base_url="",
+ agnes_api_key="agnes-key",
+ )
+ )
+
+ models = config.provider_smoke_models()
+
+ assert [model.provider for model in models] == ["agnes"]
+ assert models[0].full_model == "agnes/agnes-2.0-flash"
+ assert models[0].source == "provider_default"
+
+
def test_connected_account_provider_smoke_requires_explicit_model(
monkeypatch,
) -> None:
diff --git a/tests/providers/test_agnes.py b/tests/providers/test_agnes.py
new file mode 100644
index 00000000..5c88648d
--- /dev/null
+++ b/tests/providers/test_agnes.py
@@ -0,0 +1,227 @@
+"""Tests for the Agnes AI OpenAI-chat provider profile."""
+
+from typing import Any
+from unittest.mock import 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 AGNES_DEFAULT_BASE
+from free_claude_code.core.anthropic.models import MessagesRequest
+from free_claude_code.core.reasoning import ReasoningPolicy
+from free_claude_code.providers.base import ProviderConfig
+from free_claude_code.providers.openai_chat import OpenAIChatProvider
+from tests.providers.support import (
+ REASONING_OFF,
+ REASONING_ON,
+ immediate_admission,
+ profiled_provider,
+ reasoning_for,
+)
+
+
+def agnes_provider() -> OpenAIChatProvider:
+ return profiled_provider(
+ "agnes",
+ ProviderConfig(
+ api_key="test-agnes-key",
+ base_url=AGNES_DEFAULT_BASE,
+ rate_limit=10,
+ rate_window=60,
+ ),
+ admission=immediate_admission(provider_name="agnes"),
+ )
+
+
+def _request(**overrides: Any) -> MessagesRequest:
+ payload: dict[str, Any] = {
+ "model": "agnes-2.0-flash",
+ "messages": [{"role": "user", "content": "Hello"}],
+ }
+ payload.update(overrides)
+ return MessagesRequest.model_validate(payload)
+
+
+def test_init_uses_documented_endpoint(
+ agnes_provider: OpenAIChatProvider,
+) -> None:
+ assert agnes_provider._api_key == "test-agnes-key"
+ assert agnes_provider._base_url == AGNES_DEFAULT_BASE
+ assert agnes_provider._provider_name == "AGNES"
+
+
+def test_build_request_body_preserves_common_chat_tools_and_images(
+ agnes_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 = agnes_provider._build_request_body(
+ request,
+ reasoning=reasoning_for(request),
+ )
+
+ assert body["model"] == "agnes-2.0-flash"
+ 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"
+
+
+ ("reasoning", "enabled"),
+ [(REASONING_OFF, False), (REASONING_ON, True)],
+)
+def test_build_request_body_encodes_documented_thinking_control(
+ agnes_provider: OpenAIChatProvider,
+ reasoning: ReasoningPolicy,
+ enabled: bool,
+) -> None:
+ request = _request()
+
+ body = agnes_provider._build_request_body(request, reasoning=reasoning)
+
+ assert body["extra_body"] == {"chat_template_kwargs": {"enable_thinking": enabled}}
+
+
+def test_build_request_body_omits_thinking_control_for_provider_default(
+ agnes_provider: OpenAIChatProvider,
+) -> None:
+ request = _request()
+
+ body = agnes_provider._build_request_body(
+ request,
+ reasoning=ReasoningPolicy.provider_default(),
+ )
+
+ assert "extra_body" not in body
+
+
+def test_build_request_body_replays_reasoning_in_content_not_undocumented_field(
+ agnes_provider: OpenAIChatProvider,
+) -> None:
+ request = _request(
+ messages=[
+ {"role": "user", "content": "Solve it."},
+ {
+ "role": "assistant",
+ "content": [
+ {"type": "thinking", "thinking": "Work through it."},
+ {"type": "text", "text": "The answer is 42."},
+ ],
+ },
+ {"role": "user", "content": "Continue."},
+ ]
+ )
+
+ body = agnes_provider._build_request_body(
+ request,
+ reasoning=reasoning_for(request),
+ )
+
+ assistant = body["messages"][1]
+ assert assistant == {
+ "role": "assistant",
+ "content": "<think>\nWork through it.\n</think>\n\nThe answer is 42.",
+ }
+
+
[email protected]("field", ["reasoning", "reasoning_effort"])
+def test_build_request_body_rejects_caller_reasoning_override(
+ agnes_provider: OpenAIChatProvider,
+ field: str,
+) -> None:
+ request = _request(extra_body={field: "caller-owned"})
+
+ with pytest.raises(InvalidRequestError, match="must not override reasoning"):
+ agnes_provider._build_request_body(request, reasoning=REASONING_ON)
+
+
+async def test_model_catalog_uses_documented_endpoint_and_auth() -> None:
+ requests: list[httpx.Request] = []
+
+ def handler(request: httpx.Request) -> httpx.Response:
+ requests.append(request)
+ return httpx.Response(
+ 200,
+ json={
+ "object": "list",
+ "data": [
+ {
+ "id": "agnes-2.0-flash",
+ "object": "model",
+ "created": 0,
+ "owned_by": "agnes-ai",
+ }
+ ],
+ },
+ )
+
+ 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(
+ "agnes",
+ ProviderConfig(
+ api_key="wire-agnes-key",
+ base_url=AGNES_DEFAULT_BASE,
+ rate_limit=10,
+ rate_window=60,
+ ),
+ admission=immediate_admission(provider_name="agnes"),
+ )
+ try:
+ model_infos = await provider.list_model_infos()
+ finally:
+ await provider.cleanup()
+
+ assert model_infos == frozenset({ProviderModelInfo("agnes-2.0-flash")})
+ assert len(requests) == 1
+ assert str(requests[0].url) == "https://apihub.agnes-ai.com/v1/models"
+ assert requests[0].headers["authorization"] == "Bearer wire-agnes-key"
diff --git a/tests/providers/test_provider_runtime.py b/tests/providers/test_provider_runtime.py
index 6250e0c3..0513262e 100644
--- a/tests/providers/test_provider_runtime.py
+++ b/tests/providers/test_provider_runtime.py
@@ -11,6 +11,7 @@ from free_claude_code.application.errors import (
)
from free_claude_code.config.nim import NimSettings
from free_claude_code.config.provider_catalog import (
+ AGNES_DEFAULT_BASE,
BEDROCK_DEFAULT_BASE,
COHERE_DEFAULT_BASE,
DEEPINFRA_DEFAULT_BASE,
@@ -85,6 +86,7 @@ def _make_settings(**overrides):
mock.tokenrouter_base_url = TOKENROUTER_DEFAULT_BASE
mock.nararoute_api_key = "test_nararoute_key"
mock.nararoute_base_url = NARAROUTE_DEFAULT_BASE
+ mock.agnes_api_key = "test_agnes_key"
mock.lm_studio_base_url = "http://localhost:1234/v1"
mock.llamacpp_base_url = "http://localhost:8080/v1"
mock.ollama_base_url = "http://localhost:11434"
@@ -111,6 +113,7 @@ def _make_settings(**overrides):
mock.zai_proxy = ""
mock.tokenrouter_proxy = ""
mock.nararoute_proxy = ""
+ mock.agnes_proxy = ""
mock.fireworks_proxy = ""
mock.fireworks_api_key = "test_fireworks_key"
mock.novita_proxy = ""
@@ -280,6 +283,26 @@ def test_deepinfra_provider_config_uses_key_base_and_proxy() -> None:
assert isinstance(provider, OpenAIChatProvider)
+def test_agnes_provider_config_uses_key_base_and_proxy() -> None:
+ descriptor = PROVIDER_CATALOG["agnes"]
+ settings = _make_settings(
+ agnes_api_key="agnes-token",
+ agnes_proxy="http://proxy.test:8080",
+ )
+
+ config = build_provider_config(descriptor, settings)
+ with patch("free_claude_code.providers.openai_chat.provider.AsyncOpenAI"):
+ provider = create_provider("agnes", settings)
+
+ assert descriptor.display_name == "Agnes AI"
+ assert descriptor.credential_env == "AGNES_API_KEY"
+ assert descriptor.base_url_attr is None
+ assert config.api_key == "agnes-token"
+ assert config.base_url == AGNES_DEFAULT_BASE
+ assert config.proxy == "http://proxy.test:8080"
+ assert isinstance(provider, OpenAIChatProvider)
+
+
def test_bedrock_provider_config_uses_regional_base_key_and_proxy() -> None:
descriptor = PROVIDER_CATALOG["bedrock"]
settings = _make_settings(
@@ -601,6 +624,7 @@ def test_create_provider_instantiates_each_builtin():
"zai": OpenAIChatProvider,
"tokenrouter": OpenAIChatProvider,
"nararoute": OpenAIChatProvider,
+ "agnes": OpenAIChatProvider,
"gemini": GeminiProvider,
"vertex": VertexProvider,
"groq": GroqProvider,
diff --git a/uv.lock b/uv.lock
index 1043072f..c89c0448 100644
--- a/uv.lock
+++ b/uv.lock
@@ -620,7 +620,7 @@ wheels = [
[[package]]
name = "free-claude-code"
-version = "4.24.0"
+version = "4.25.0"
source = { editable = "." }
dependencies = [
{ name = "aiohttp" },