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 SambaNova Cloud provider (#990)

newmemories360770d56708aa4
diff --git a/.env.example b/.env.example
index c7839929..2ab2f50c 100644
--- a/.env.example
+++ b/.env.example
@@ -71,6 +71,10 @@ GEMINI_API_KEY=""
GROQ_API_KEY=""
+# SambaNova Cloud (OpenAI-compatible Chat Completions at api.sambanova.ai/v1)
+SAMBANOVA_API_KEY=""
+
+
# Cerebras Inference (OpenAI-compatible Chat Completions; see https://inference-docs.cerebras.ai/resources/openai)
CEREBRAS_API_KEY=""
@@ -89,7 +93,7 @@ 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" | "open_router" | "gemini" | "deepseek" | "mistral" | "mistral_codestral" | "opencode" | "opencode_go" | "vercel" | "huggingface" | "cohere" | "github_models" | "wafer" | "kimi" | "minimax" | "cerebras" | "groq" | "fireworks" | "cloudflare" | "zai" | "lmstudio" | "llamacpp" | "ollama"
+# Valid providers: "nvidia_nim" | "open_router" | "gemini" | "deepseek" | "mistral" | "mistral_codestral" | "opencode" | "opencode_go" | "vercel" | "huggingface" | "cohere" | "github_models" | "wafer" | "kimi" | "minimax" | "cerebras" | "groq" | "sambanova" | "fireworks" | "cloudflare" | "zai" | "lmstudio" | "llamacpp" | "ollama"
MODEL_OPUS=
MODEL_SONNET=
MODEL_HAIKU=
@@ -120,6 +124,7 @@ FCC_SMOKE_MODEL_FIREWORKS=
FCC_SMOKE_MODEL_CLOUDFLARE=
FCC_SMOKE_MODEL_GEMINI=
FCC_SMOKE_MODEL_GROQ=
+FCC_SMOKE_MODEL_SAMBANOVA=
FCC_SMOKE_MODEL_CEREBRAS=
FCC_SMOKE_NIM_MODELS=
FCC_SMOKE_NIM_EXTRA_MODELS=
@@ -158,6 +163,7 @@ FIREWORKS_PROXY=""
CLOUDFLARE_PROXY=""
GEMINI_PROXY=""
GROQ_PROXY=""
+SAMBANOVA_PROXY=""
CEREBRAS_PROXY=""
PROVIDER_RATE_LIMIT=1
diff --git a/README.md b/README.md
index 9760d341..a5e577e9 100644
--- a/README.md
+++ b/README.md
@@ -57,7 +57,7 @@ Free Claude Code routes Anthropic Messages API traffic from Claude Code (CLI and
- Drop-in proxy for Claude Code's Anthropic API calls (`/v1/messages`, `/v1/models`).
- Drop-in proxy for Codex via the OpenAI Responses API (`/v1/responses`).
- `fcc-claude` and `fcc-codex` launchers that read the current Admin UI port and auth token each time they start.
-- 23 provider backends: NVIDIA NIM, OpenRouter, Google AI Studio (Gemini), DeepSeek, Mistral La Plateforme, Mistral Codestral, OpenCode Zen, OpenCode Go, Vercel AI Gateway, Hugging Face Inference Providers, Cohere, GitHub Models, Wafer, Kimi, MiniMax, Cerebras Inference, Groq, Fireworks AI, Cloudflare, Z.ai, LM Studio, llama.cpp, and Ollama.
+- 24 provider backends: NVIDIA NIM, OpenRouter, Google AI Studio (Gemini), DeepSeek, Mistral La Plateforme, Mistral Codestral, OpenCode Zen, OpenCode Go, Vercel AI Gateway, Hugging Face Inference Providers, Cohere, GitHub Models, Wafer, Kimi, MiniMax, Cerebras Inference, Groq, SambaNova, Fireworks AI, Cloudflare, Z.ai, LM Studio, llama.cpp, and Ollama.
- Per-model routing for Claude Code: send Opus, Sonnet, Haiku, and fallback traffic to different providers.
- Native Claude Code `/model` picker support through the proxy's `/v1/models` endpoint (see [Model Picker](#model-picker)).
- Native Codex `/model` picker support when launched through `fcc-codex`, using a generated local model catalog.
@@ -346,7 +346,17 @@ Reasoning-heavy models expose extra knobs documented under [Groq reasoning](http
Browse models at [console.groq.com/docs/models](https://console.groq.com/docs/models).
-### 18. [Fireworks AI](https://fireworks.ai/)
+### 18. [SambaNova](https://sambanova.ai/)
+
+Create an API key in the [SambaNova Cloud console](https://cloud.sambanova.ai/apis).
+
+In the Admin UI, paste it into `SAMBANOVA_API_KEY`, then set `MODEL` to a SambaNova model slug such as `sambanova/Meta-Llama-3.3-70B-Instruct`.
+
+SambaNova Cloud exposes an OpenAI-compatible Chat Completions API at `https://api.sambanova.ai/v1` ([OpenAI compatibility](https://docs.sambanova.ai/cloud/docs/capabilities/openai-compatibility)). This proxy follows the shared OpenAI-compat adapters, mapping Claude-style thinking to `reasoning_content` deltas where the model supports it.
+
+Browse models at [SambaNova Cloud models](https://docs.sambanova.ai/cloud/docs/get-started/supported-models).
+
+### 19. [Fireworks AI](https://fireworks.ai/)
Get an API key at [fireworks.ai/account/api-keys](https://fireworks.ai/account/api-keys).
@@ -356,7 +366,7 @@ Fireworks exposes an **Anthropic-compatible** Messages API at `https://api.firew
Browse models at [fireworks.ai/models](https://fireworks.ai/models).
-### 19. [Cloudflare](https://developers.cloudflare.com/workers-ai/)
+### 20. [Cloudflare](https://developers.cloudflare.com/workers-ai/)
Create a Cloudflare API token and copy your account ID from the Cloudflare dashboard.
@@ -364,7 +374,7 @@ In the Admin UI, set `CLOUDFLARE_API_TOKEN` and `CLOUDFLARE_ACCOUNT_ID`, then se
This provider calls Cloudflare's account-scoped **OpenAI-compatible** Chat Completions API at `https://api.cloudflare.com/client/v4/accounts/<account_id>/ai/v1/chat/completions`. Use literal Workers AI model IDs, including the `@cf/` prefix when the catalog model includes it.
-### 20. [Z.ai](https://z.ai/)
+### 21. [Z.ai](https://z.ai/)
Get an API key at [Z.ai/manage-apikey/apikey-list](https://z.ai/manage-apikey/apikey-list).
@@ -379,13 +389,13 @@ Popular examples:
Browse models at [Z.ai](https://z.ai).
-### 21. [LM Studio](https://lmstudio.ai/)
+### 22. [LM Studio](https://lmstudio.ai/)
Start LM Studio's local server and load a model. In the Admin UI, keep or update `LM_STUDIO_BASE_URL`, then set `MODEL` to the model identifier shown by LM Studio, prefixed with `lmstudio/`.
Prefer models with tool-use support for Claude Code workflows.
-### 22. [llama.cpp](https://github.com/ggml-org/llama.cpp)
+### 23. [llama.cpp](https://github.com/ggml-org/llama.cpp)
Start `llama-server` with an Anthropic-compatible `/v1/messages` endpoint and enough context for Claude Code requests.
@@ -393,7 +403,7 @@ In the Admin UI, keep or update `LLAMACPP_BASE_URL`, then set `MODEL` to the loc
For local coding models, context size matters. If llama.cpp returns HTTP 400 for normal Claude Code requests, increase `--ctx-size` and verify the model/server build supports the requested features.
-### 23. [Ollama](https://ollama.com/)
+### 24. [Ollama](https://ollama.com/)
Run Ollama and pull a model:
@@ -406,7 +416,7 @@ In the Admin UI, keep or update `OLLAMA_BASE_URL`, then set `MODEL` to the same
`OLLAMA_BASE_URL` is the Ollama server root; do not append `/v1`. Example model slugs include `ollama/llama3.1` and `ollama/llama3.1:8b`.
-### 24. Mix Providers By Model Tier
+### 25. Mix Providers By Model Tier
Each model tier can use a different provider by setting `MODEL_OPUS`, `MODEL_SONNET`, and `MODEL_HAIKU` in the Admin UI. Leave a tier blank to inherit `MODEL`. These tier overrides apply to Claude model names that contain `opus`, `sonnet`, or `haiku`. Codex uses the Admin `MODEL` default through `fcc-codex` unless a session requests a provider-prefixed slug directly.
@@ -620,7 +630,7 @@ Important pieces:
- Responses requests convert to Anthropic Messages internally, then share the same model router, normalizer, and provider adapters.
- `fcc-codex` registers a custom `fcc` provider that points Codex at the local proxy's `/v1/responses` endpoint.
- Model routing resolves Claude model names to `MODEL_OPUS`, `MODEL_SONNET`, `MODEL_HAIKU`, or `MODEL`.
-- NIM, Gemini, DeepSeek, Mistral, Codestral, OpenCode Zen, OpenCode Go, Vercel AI Gateway, Hugging Face, Cohere, GitHub Models, Cerebras, Groq, and Cloudflare use OpenAI chat streaming translated into Anthropic SSE.
+- NIM, Gemini, DeepSeek, Mistral, Codestral, OpenCode Zen, OpenCode Go, Vercel AI Gateway, Hugging Face, Cohere, GitHub Models, Cerebras, Groq, SambaNova, and Cloudflare use OpenAI chat streaming translated into Anthropic SSE.
- Wafer, OpenRouter, Kimi, MiniMax, Fireworks AI, Z.ai, LM Studio, llama.cpp, and Ollama use Anthropic Messages style transports where applicable (with provider-specific quirks and model-list URLs).
- The proxy normalizes thinking blocks, tool calls, token usage metadata, and provider errors into the shape each client expects.
- Request optimizations answer trivial Claude Code probes locally to save latency and quota.
diff --git a/api/admin_config/manifest.py b/api/admin_config/manifest.py
index 34a18006..4ba0a753 100644
--- a/api/admin_config/manifest.py
+++ b/api/admin_config/manifest.py
@@ -610,6 +610,12 @@ _NON_PROVIDER_FIELDS: tuple[ConfigFieldSpec, ...] = (
"smoke",
advanced=True,
),
+ ConfigFieldSpec(
+ "FCC_SMOKE_MODEL_SAMBANOVA",
+ "Smoke SambaNova Model",
+ "smoke",
+ advanced=True,
+ ),
ConfigFieldSpec(
"FCC_SMOKE_MODEL_CEREBRAS",
"Smoke Cerebras Model",
diff --git a/api/admin_config/provider_manifest.py b/api/admin_config/provider_manifest.py
index 47b5f826..f0282d49 100644
--- a/api/admin_config/provider_manifest.py
+++ b/api/admin_config/provider_manifest.py
@@ -94,6 +94,13 @@ _PROVIDER_FIELD_OVERRIDES: dict[str, dict[str, Any]] = {
"[OpenAI compatibility docs](https://console.groq.com/docs/openai)."
),
},
+ "SAMBANOVA_API_KEY": {
+ "label": "SambaNova API Key",
+ "description": (
+ "SambaNova Cloud OpenAI-compatible API key (create at "
+ "[cloud.sambanova.ai/apis](https://cloud.sambanova.ai/apis))."
+ ),
+ },
"CEREBRAS_API_KEY": {
"label": "Cerebras API Key",
"description": (
diff --git a/config/provider_catalog.py b/config/provider_catalog.py
index 80ff41f5..8d2c248e 100644
--- a/config/provider_catalog.py
+++ b/config/provider_catalog.py
@@ -39,6 +39,7 @@ ZAI_DEFAULT_BASE = "https://api.z.ai/api/anthropic/v1"
GEMINI_DEFAULT_BASE = "https://generativelanguage.googleapis.com/v1beta/openai/"
GROQ_DEFAULT_BASE = "https://api.groq.com/openai/v1"
CEREBRAS_DEFAULT_BASE = "https://api.cerebras.ai/v1"
+SAMBANOVA_DEFAULT_BASE = "https://api.sambanova.ai/v1"
@dataclass(frozen=True, slots=True)
@@ -258,6 +259,17 @@ PROVIDER_CATALOG: dict[str, ProviderDescriptor] = {
proxy_attr="groq_proxy",
capabilities=("chat", "streaming", "tools", "thinking", "rate_limit"),
),
+ "sambanova": ProviderDescriptor(
+ provider_id="sambanova",
+ display_name="SambaNova",
+ transport_type="openai_chat",
+ credential_env="SAMBANOVA_API_KEY",
+ credential_url="https://cloud.sambanova.ai/apis",
+ credential_attr="sambanova_api_key",
+ default_base_url=SAMBANOVA_DEFAULT_BASE,
+ proxy_attr="sambanova_proxy",
+ capabilities=("chat", "streaming", "tools", "thinking", "rate_limit"),
+ ),
"fireworks": ProviderDescriptor(
provider_id="fireworks",
display_name="Fireworks",
diff --git a/config/settings.py b/config/settings.py
index ef40abaa..a47e6b8a 100644
--- a/config/settings.py
+++ b/config/settings.py
@@ -58,6 +58,9 @@ class Settings(BaseSettings):
# ==================== GitHub Models ====================
github_models_token: str = Field(default="", validation_alias="GITHUB_MODELS_TOKEN")
+ # ==================== SambaNova Cloud ====================
+ sambanova_api_key: str = Field(default="", validation_alias="SAMBANOVA_API_KEY")
+
# ==================== Z.ai Config ====================
zai_api_key: str = Field(default="", validation_alias="ZAI_API_KEY")
@@ -143,6 +146,7 @@ class Settings(BaseSettings):
huggingface_proxy: str = Field(default="", validation_alias="HUGGINGFACE_PROXY")
cohere_proxy: str = Field(default="", validation_alias="COHERE_PROXY")
github_models_proxy: str = Field(default="", validation_alias="GITHUB_MODELS_PROXY")
+ sambanova_proxy: str = Field(default="", validation_alias="SAMBANOVA_PROXY")
zai_proxy: str = Field(default="", validation_alias="ZAI_PROXY")
fireworks_proxy: str = Field(default="", validation_alias="FIREWORKS_PROXY")
cloudflare_proxy: str = Field(default="", validation_alias="CLOUDFLARE_PROXY")
diff --git a/providers/defaults.py b/providers/defaults.py
index b70773e7..2c8a79ba 100644
--- a/providers/defaults.py
+++ b/providers/defaults.py
@@ -20,6 +20,7 @@ from config.provider_catalog import (
OPENCODE_DEFAULT_BASE,
OPENCODE_GO_DEFAULT_BASE,
OPENROUTER_DEFAULT_BASE,
+ SAMBANOVA_DEFAULT_BASE,
VERCEL_AI_GATEWAY_DEFAULT_BASE,
WAFER_DEFAULT_BASE,
ZAI_DEFAULT_BASE,
@@ -45,6 +46,7 @@ __all__ = (
"OPENCODE_DEFAULT_BASE",
"OPENCODE_GO_DEFAULT_BASE",
"OPENROUTER_DEFAULT_BASE",
+ "SAMBANOVA_DEFAULT_BASE",
"VERCEL_AI_GATEWAY_DEFAULT_BASE",
"WAFER_DEFAULT_BASE",
"ZAI_DEFAULT_BASE",
diff --git a/providers/runtime/factory.py b/providers/runtime/factory.py
index 7ee84da6..472ef97a 100644
--- a/providers/runtime/factory.py
+++ b/providers/runtime/factory.py
@@ -149,6 +149,12 @@ def _create_groq(config: ProviderConfig, _settings: Settings) -> BaseProvider:
return GroqProvider(config)
+def _create_sambanova(config: ProviderConfig, _settings: Settings) -> BaseProvider:
+ from providers.sambanova import SambaNovaProvider
+
+ return SambaNovaProvider(config)
+
+
def _create_cerebras(config: ProviderConfig, _settings: Settings) -> BaseProvider:
from providers.cerebras import CerebrasProvider
@@ -173,6 +179,7 @@ PROVIDER_FACTORIES: dict[str, ProviderFactory] = {
"minimax": _create_minimax,
"cerebras": _create_cerebras,
"groq": _create_groq,
+ "sambanova": _create_sambanova,
"fireworks": _create_fireworks,
"cloudflare": _create_cloudflare,
"zai": _create_zai,
diff --git a/providers/sambanova/__init__.py b/providers/sambanova/__init__.py
new file mode 100644
index 00000000..6ff65ced
--- /dev/null
+++ b/providers/sambanova/__init__.py
@@ -0,0 +1,7 @@
+"""SambaNova Cloud OpenAI-compatible adapter."""
+
+from providers.defaults import SAMBANOVA_DEFAULT_BASE
+
+from .client import SambaNovaProvider
+
+__all__ = ["SAMBANOVA_DEFAULT_BASE", "SambaNovaProvider"]
diff --git a/providers/sambanova/client.py b/providers/sambanova/client.py
new file mode 100644
index 00000000..da403a82
--- /dev/null
+++ b/providers/sambanova/client.py
@@ -0,0 +1,37 @@
+"""SambaNova Cloud provider implementation (OpenAI-compatible chat completions)."""
+
+from typing import Any
+
+from providers.base import ProviderConfig
+from providers.defaults import SAMBANOVA_DEFAULT_BASE
+from providers.transports.openai_chat import (
+ OpenAIChatRequestPolicy,
+ OpenAIChatTransport,
+ build_openai_chat_request_body,
+)
+
+_REQUEST_POLICY = OpenAIChatRequestPolicy(
+ provider_name="SAMBANOVA",
+ include_extra_body=True,
+)
+
+
+class SambaNovaProvider(OpenAIChatTransport):
+ """SambaNova Cloud API at ``https://api.sambanova.ai/v1``."""
+
+ def __init__(self, config: ProviderConfig):
+ super().__init__(
+ config,
+ provider_name="SAMBANOVA",
+ base_url=config.base_url or SAMBANOVA_DEFAULT_BASE,
+ api_key=config.api_key,
+ )
+
+ def _build_request_body(
+ self, request: Any, thinking_enabled: bool | None = None
+ ) -> dict:
+ return build_openai_chat_request_body(
+ request,
+ thinking_enabled=self._is_thinking_enabled(request, thinking_enabled),
+ policy=_REQUEST_POLICY,
+ )
diff --git a/pyproject.toml b/pyproject.toml
index 999ff2fa..2dd35785 100644
--- a/pyproject.toml
+++ b/pyproject.toml
@@ -4,7 +4,7 @@ build-backend = "hatchling.build"
[project]
name = "free-claude-code"
-version = "3.2.1"
+version = "3.3.0"
description = "Middleware between Claude Code CLI (Anthropic API) and NVIDIA NIM"
readme = "README.md"
requires-python = ">=3.14.0"
diff --git a/smoke/lib/config.py b/smoke/lib/config.py
index 6aa299e0..b68e4144 100644
--- a/smoke/lib/config.py
+++ b/smoke/lib/config.py
@@ -60,6 +60,7 @@ PROVIDER_SMOKE_DEFAULT_MODELS: dict[str, str] = {
"zai": "zai/glm-5.1",
"gemini": "gemini/models/gemini-3.1-flash-lite",
"groq": "groq/llama-3.3-70b-versatile",
+ "sambanova": "sambanova/Meta-Llama-3.3-70B-Instruct",
"cerebras": "cerebras/llama3.1-8b",
"cloudflare": "cloudflare/@cf/moonshotai/kimi-k2.6",
}
@@ -270,6 +271,8 @@ class SmokeConfig:
return bool(self.settings.gemini_api_key.strip())
if provider == "groq":
return bool(self.settings.groq_api_key.strip())
+ if provider == "sambanova":
+ return bool(self.settings.sambanova_api_key.strip())
if provider == "cerebras":
return bool(self.settings.cerebras_api_key.strip())
if provider == "cloudflare":
diff --git a/tests/api/test_admin.py b/tests/api/test_admin.py
index 1f9aafa6..0938aa43 100644
--- a/tests/api/test_admin.py
+++ b/tests/api/test_admin.py
@@ -30,6 +30,7 @@ def _clear_process_config(monkeypatch) -> None:
"CLOUDFLARE_API_TOKEN",
"CLOUDFLARE_ACCOUNT_ID",
"GITHUB_MODELS_TOKEN",
+ "SAMBANOVA_API_KEY",
"HOST",
"PORT",
"LOG_FILE",
@@ -109,6 +110,7 @@ def test_admin_config_masks_secrets_and_exposes_manifest(monkeypatch, tmp_path):
assert "GITHUB_MODELS_TOKEN" in keys
assert "GEMINI_API_KEY" in keys
assert "GROQ_API_KEY" in keys
+ assert "SAMBANOVA_API_KEY" in keys
assert "CEREBRAS_API_KEY" in keys
assert "ZAI_BASE_URL" not in keys
assert "CLAUDE_WORKSPACE" not in keys
@@ -264,6 +266,31 @@ def test_admin_apply_writes_groq_key_and_masks_preview(monkeypatch, tmp_path):
assert "GROQ_API_KEY=gq-secret" in text
+def test_admin_apply_writes_sambanova_key_and_masks_preview(monkeypatch, tmp_path):
+ _set_home(monkeypatch, tmp_path)
+ _clear_process_config(monkeypatch)
+ app = create_app(lifespan_enabled=False)
+
+ response = _local_client(app).post(
+ "/admin/api/config/apply",
+ json={
+ "values": {
+ "MODEL": "sambanova/Meta-Llama-3.3-70B-Instruct",
+ "SAMBANOVA_API_KEY": "sn-secret",
+ }
+ },
+ )
+
+ assert response.status_code == 200
+ body = response.json()
+ assert body["applied"] is True
+ assert "SAMBANOVA_API_KEY=********" in body["env_preview"]
+ env_file = tmp_path / ".fcc" / ".env"
+ text = env_file.read_text(encoding="utf-8")
+ assert "MODEL=sambanova/Meta-Llama-3.3-70B-Instruct" in text
+ assert "SAMBANOVA_API_KEY=sn-secret" in text
+
+
def test_admin_apply_writes_cerebras_key_and_masks_preview(monkeypatch, tmp_path):
_set_home(monkeypatch, tmp_path)
_clear_process_config(monkeypatch)
diff --git a/tests/config/test_config.py b/tests/config/test_config.py
index 01669fff..fd77642c 100644
--- a/tests/config/test_config.py
+++ b/tests/config/test_config.py
@@ -347,6 +347,16 @@ class TestSettings:
assert settings.github_models_token == "github-token"
assert settings.github_models_proxy == "http://proxy.test:8080"
+ def test_sambanova_settings_from_env(self, monkeypatch):
+ """SambaNova key and proxy env vars load into settings."""
+ from config.settings import Settings
+
+ monkeypatch.setenv("SAMBANOVA_API_KEY", "sambanova-key")
+ monkeypatch.setenv("SAMBANOVA_PROXY", "http://proxy.test:8080")
+ settings = Settings()
+ assert settings.sambanova_api_key == "sambanova-key"
+ assert settings.sambanova_proxy == "http://proxy.test:8080"
+
def test_legacy_hf_token_env_is_ignored(self, monkeypatch):
"""HF_TOKEN is migrated by startup config migration, not read by Settings."""
from config.settings import Settings
@@ -754,6 +764,11 @@ class TestPerModelMapping:
"github_models/openai/gpt-4.1",
None,
),
+ (
+ {"MODEL": "sambanova/Meta-Llama-3.3-70B-Instruct"},
+ "sambanova/Meta-Llama-3.3-70B-Instruct",
+ None,
+ ),
({"MODEL": "lmstudio/qwen2.5-7b"}, "lmstudio/qwen2.5-7b", None),
({"MODEL": "llamacpp/local-model"}, "llamacpp/local-model", None),
({"MODEL": "ollama/llama3.1"}, "ollama/llama3.1", None),
@@ -948,6 +963,9 @@ class TestPerModelMapping:
assert parse_provider_type("github_models/openai/gpt-4.1") == ("github_models")
assert parse_provider_type("gemini/models/gemini-3.1-flash-lite") == "gemini"
assert parse_provider_type("groq/llama-3.3-70b-versatile") == "groq"
+ assert (
+ parse_provider_type("sambanova/Meta-Llama-3.3-70B-Instruct") == "sambanova"
+ )
assert parse_provider_type("cerebras/llama3.1-8b") == "cerebras"
def test_parse_model_name(self):
@@ -987,6 +1005,10 @@ class TestPerModelMapping:
parse_model_name("groq/llama-3.3-70b-versatile")
== "llama-3.3-70b-versatile"
)
+ assert (
+ parse_model_name("sambanova/Meta-Llama-3.3-70B-Instruct")
+ == "Meta-Llama-3.3-70B-Instruct"
+ )
assert parse_model_name("cerebras/llama3.1-8b") == "llama3.1-8b"
def test_configured_chat_model_refs_collects_unique_models_with_sources(
diff --git a/tests/contracts/test_feature_manifest.py b/tests/contracts/test_feature_manifest.py
index 6b1b09f2..5fe344a7 100644
--- a/tests/contracts/test_feature_manifest.py
+++ b/tests/contracts/test_feature_manifest.py
@@ -22,6 +22,7 @@ from providers.nvidia_nim import NvidiaNimProvider
from providers.ollama import OllamaProvider
from providers.open_router import OpenRouterProvider
from providers.opencode import OpenCodeProvider
+from providers.sambanova import SambaNovaProvider
from providers.vercel import VercelProvider
from providers.wafer import WaferProvider
from providers.zai import ZaiProvider
@@ -103,6 +104,7 @@ def test_provider_and_platform_registries_include_advertised_builtins() -> None:
"zai": ZaiProvider,
"gemini": GeminiProvider,
"groq": GroqProvider,
+ "sambanova": SambaNovaProvider,
"cerebras": CerebrasProvider,
}
for provider_class in provider_classes.values():
diff --git a/tests/contracts/test_provider_catalog_order.py b/tests/contracts/test_provider_catalog_order.py
index dcba6556..1cdb1e71 100644
--- a/tests/contracts/test_provider_catalog_order.py
+++ b/tests/contracts/test_provider_catalog_order.py
@@ -20,6 +20,7 @@ _EXPECTED_PROVIDER_ORDER: tuple[str, ...] = (
"minimax",
"cerebras",
"groq",
+ "sambanova",
"fireworks",
"cloudflare",
"zai",
diff --git a/tests/contracts/test_smoke_config.py b/tests/contracts/test_smoke_config.py
index b375bb35..80137c8c 100644
--- a/tests/contracts/test_smoke_config.py
+++ b/tests/contracts/test_smoke_config.py
@@ -42,6 +42,7 @@ def _settings(**overrides):
"zai_api_key": "",
"gemini_api_key": "",
"groq_api_key": "",
+ "sambanova_api_key": "",
"cerebras_api_key": "",
"fireworks_api_key": "",
"cloudflare_api_token": "",
@@ -260,6 +261,22 @@ def test_github_models_provider_configuration_uses_token(monkeypatch) -> None:
assert models[0].full_model == PROVIDER_SMOKE_DEFAULT_MODELS["github_models"]
+def test_sambanova_provider_configuration_uses_api_key(monkeypatch) -> None:
+ monkeypatch.delenv("FCC_SMOKE_MODEL_SAMBANOVA", raising=False)
+ config = _smoke_config(
+ settings=_settings(
+ model="ollama/llama3.1",
+ ollama_base_url="",
+ sambanova_api_key="sambanova-key",
+ )
+ )
+
+ assert config.has_provider_configuration("sambanova")
+ models = config.provider_smoke_models()
+ assert models[0].provider == "sambanova"
+ assert models[0].full_model == PROVIDER_SMOKE_DEFAULT_MODELS["sambanova"]
+
+
def test_provider_smoke_model_override_accepts_model_name_without_prefix(
monkeypatch,
) -> None:
diff --git a/tests/providers/test_sambanova.py b/tests/providers/test_sambanova.py
new file mode 100644
index 00000000..0577ce41
--- /dev/null
+++ b/tests/providers/test_sambanova.py
@@ -0,0 +1,224 @@
+"""Tests for SambaNova Cloud (OpenAI-compatible) provider."""
+
+from contextlib import asynccontextmanager
+from unittest.mock import AsyncMock, MagicMock, patch
+
+import pytest
+
+from providers.base import ProviderConfig
+from providers.sambanova import SAMBANOVA_DEFAULT_BASE, SambaNovaProvider
+
+
+class MockMessage:
+ def __init__(self, role, content):
+ self.role = role
+ self.content = content
+
+
+class MockRequest:
+ def __init__(self, **kwargs):
+ self.model = "Meta-Llama-3.3-70B-Instruct"
+ self.messages = [MockMessage("user", "Hello")]
+ self.max_tokens = 100
+ self.temperature = 0.5
+ self.top_p = 0.9
+ self.system = "System prompt"
+ self.stop_sequences = None
+ self.tools = []
+ self.thinking = MagicMock()
+ self.thinking.enabled = True
+ for key, value in kwargs.items():
+ setattr(self, key, value)
+
+
+def sambanova_config():
+ return ProviderConfig(
+ api_key="test_sambanova_key",
+ base_url=SAMBANOVA_DEFAULT_BASE,
+ rate_limit=10,
+ rate_window=60,
+ enable_thinking=True,
+ )
+
+
[email protected](autouse=True)
+def mock_rate_limiter():
+ """Mock the global rate limiter to prevent waiting."""
+
+ @asynccontextmanager
+ async def _slot():
+ yield
+
+ with patch("providers.transports.openai_chat.transport.GlobalRateLimiter") as mock:
+ instance = mock.get_scoped_instance.return_value
+
+ async def _passthrough(fn, *args, **kwargs):
+ return await fn(*args, **kwargs)
+
+ instance.execute_with_retry = AsyncMock(side_effect=_passthrough)
+ instance.concurrency_slot.side_effect = _slot
+ yield instance
+
+
+def sambanova_provider(sambanova_config):
+ return SambaNovaProvider(sambanova_config)
+
+
+def test_default_base_url_constant():
+ assert SAMBANOVA_DEFAULT_BASE == "https://api.sambanova.ai/v1"
+
+
+def test_init_uses_default_base_url_and_api_key(sambanova_config):
+ with patch("providers.transports.openai_chat.transport.AsyncOpenAI") as mock_openai:
+ provider = SambaNovaProvider(sambanova_config)
+
+ assert provider._api_key == "test_sambanova_key"
+ assert provider._base_url == SAMBANOVA_DEFAULT_BASE
+ mock_openai.assert_called_once()
+
+
+def test_init_strips_trailing_slash(sambanova_config):
+ config = sambanova_config.model_copy(
+ update={"base_url": f"{SAMBANOVA_DEFAULT_BASE}/"}
+ )
+
+ with patch("providers.transports.openai_chat.transport.AsyncOpenAI"):
+ provider = SambaNovaProvider(config)
+
+ assert provider._base_url == SAMBANOVA_DEFAULT_BASE
+
+
+def test_build_request_body_basic(sambanova_provider):
+ """Basic request body conversion attaches system message and keeps max_tokens."""
+ body = sambanova_provider._build_request_body(MockRequest())
+
+ assert body["model"] == "Meta-Llama-3.3-70B-Instruct"
+ assert body["messages"][0]["role"] == "system"
+ assert body["max_tokens"] == 100
+ assert "max_completion_tokens" not in body
+
+
+def test_build_request_body_preserves_caller_extra_body(sambanova_provider):
+ req = MockRequest(extra_body={"metadata": {"user": "u1"}})
+
+ body = sambanova_provider._build_request_body(req)
+
+ eb = body.get("extra_body")
+ assert isinstance(eb, dict)
+ assert eb.get("metadata") == {"user": "u1"}
+
+
+async def test_stream_response_text(sambanova_provider):
+ """Text content deltas are emitted as text blocks."""
+ mock_chunk = MagicMock()
+ mock_chunk.choices = [
+ MagicMock(
+ delta=MagicMock(
+ content="Hello from SambaNova",
+ reasoning_content=None,
+ tool_calls=None,
+ ),
+ finish_reason="stop",
+ )
+ ]
+ mock_chunk.usage = MagicMock(completion_tokens=5, prompt_tokens=10)
+
+ async def mock_stream():
+ yield mock_chunk
+
+ with patch.object(
+ sambanova_provider._client.chat.completions, "create", new_callable=AsyncMock
+ ) as mock_create:
+ mock_create.return_value = mock_stream()
+
+ events = [
+ event async for event in sambanova_provider.stream_response(MockRequest())
+ ]
+
+ assert any(
+ '"text_delta"' in event and "Hello from SambaNova" in event for event in events
+ )
+
+
+async def test_stream_response_tool_call(sambanova_provider):
+ mock_tc = MagicMock()
+ mock_tc.index = 0
+ mock_tc.id = "call_1"
+ mock_tc.function = MagicMock()
+ mock_tc.function.name = "Read"
+ mock_tc.function.arguments = '{"file_path":"a.py"}'
+
+ mock_chunk = MagicMock()
+ mock_chunk.choices = [
+ MagicMock(
+ delta=MagicMock(content=None, reasoning_content=None, tool_calls=[mock_tc]),
+ finish_reason="tool_calls",
+ )
+ ]
+ mock_chunk.usage = MagicMock(completion_tokens=5, prompt_tokens=10)
+
+ async def mock_stream():
+ yield mock_chunk
+
+ with patch.object(
+ sambanova_provider._client.chat.completions, "create", new_callable=AsyncMock
+ ) as mock_create:
+ mock_create.return_value = mock_stream()
+
+ events = [
+ event async for event in sambanova_provider.stream_response(MockRequest())
+ ]
+
+ assert any(
+ '"content_block_start"' in event and '"tool_use"' in event for event in events
+ )
+ assert any(
+ '"input_json_delta"' in event and "file_path" in event for event in events
+ )
+
+
+async def test_stream_response_reasoning_content(sambanova_provider):
+ """reasoning_content deltas are emitted as thinking blocks."""
+ mock_chunk = MagicMock()
+ mock_chunk.choices = [
+ MagicMock(
+ delta=MagicMock(
+ content=None,
+ reasoning_content="Thinking via SambaNova",
+ tool_calls=None,
+ ),
+ finish_reason="stop",
+ )
+ ]
+ mock_chunk.usage = MagicMock(completion_tokens=2, prompt_tokens=10)
+
+ async def mock_stream():
+ yield mock_chunk
+
+ with patch.object(
+ sambanova_provider._client.chat.completions, "create", new_callable=AsyncMock
+ ) as mock_create:
+ mock_create.return_value = mock_stream()
+
+ events = [
+ event async for event in sambanova_provider.stream_response(MockRequest())
+ ]
+
+ assert any(
+ '"thinking_delta"' in event and "Thinking via SambaNova" in event
+ for event in events
+ )
+
+
+async def test_cleanup(sambanova_provider):
+ sambanova_provider._client = AsyncMock()
+
+ await sambanova_provider.cleanup()
+
+ sambanova_provider._client.close.assert_called_once()
diff --git a/uv.lock b/uv.lock
index 1358a9a9..3ac661e6 100644
--- a/uv.lock
+++ b/uv.lock
@@ -561,7 +561,7 @@ wheels = [
[[package]]
name = "free-claude-code"
-version = "3.2.1"
+version = "3.3.0"
source = { editable = "." }
dependencies = [
{ name = "aiohttp" },