BerriAI/litellm · error · ValueError

No provider mapping found for model {model}

Error message

No provider mapping found for model {model}

What it means

Raised in _fetch_inference_provider_mapping (litellm/llms/huggingface/common_utils.py) when the HF Hub API responds successfully for /api/models/<model> but the returned JSON has no inferenceProviderMapping field. This means the model exists on the Hub but has no inference-provider deployment mapping, so LiteLLM cannot resolve where to send the request.

Source

Thrown at litellm/llms/huggingface/common_utils.py:88

    Raises:
        ValueError: If no provider mapping is found
        HuggingFaceError: If the API request fails
    """
    headers = {"Accept": "application/json"}
    if os.getenv("HUGGINGFACE_API_KEY"):
        headers["Authorization"] = f"Bearer {os.getenv('HUGGINGFACE_API_KEY')}"

    path: Final = f"{HF_HUB_URL}/api/models/{model}"
    params: Final = {"expand": ["inferenceProviderMapping"]}

    try:
        response: Final = httpx.get(path, headers=headers, params=params)
        response.raise_for_status()
        provider_mapping: Final = response.json().get("inferenceProviderMapping")

        if provider_mapping is None:
            raise ValueError(f"No provider mapping found for model {model}")

        return provider_mapping
    except httpx.HTTPError as e:
        if hasattr(e, "response"):
            status_code = getattr(e.response, "status_code", 500)
            headers = getattr(e.response, "headers", {})
        else:
            status_code = 500
            headers = {}
        raise HuggingFaceError(
            message=f"Failed to fetch provider mapping: {e}",
            status_code=status_code,
            headers=headers,
        )

View on GitHub (pinned to 6c2dcb801b)

Solutions

  1. Open the model page on huggingface.co and check the Inference Providers tab — if empty, no provider serves it; pick a model that is deployed (or the same model under an org that deployed it).
  2. If it must be this model, self-host it (vLLM/TEI) and call it via hosted_vllm/custom_openai with api_base, or use a provider that hosts it.
  3. Double-check the model id (org/name spelling) and that it is a model repo, not a dataset/space.
  4. If the model should have providers (per the UI), it may be caching — retry after a short delay or verify with the expand=inferenceProviderMapping API call by hand.

Example fix

# before
litellm.completion(model='hf/some-org/model-without-providers', messages=msgs)
# ValueError: No provider mapping found for model some-org/model-without-providers

# after — pick a deployed model or self-host
litellm.completion(model='hf/meta-llama/Llama-3.1-8B-Instruct', messages=msgs)
# or self-host: litellm.completion(model='hosted_vllm/some-org/model-without-providers',
#                                 api_base='http://my-vllm:8000', messages=msgs)
Defensive patterns

Strategy: validation

Validate before calling

import httpx

def has_inference_providers(model_id: str) -> bool:
    r = httpx.get(f"https://huggingface.co/api/models/{model_id}",
                  params={"expand": ["inferenceProviderMapping"]}, timeout=10)
    r.raise_for_status()
    return bool(r.json().get("inferenceProviderMapping"))

if not has_inference_providers("some-org/some-model"):
    raise ValueError("model has no inference providers — pick a deployed model or self-host")

Try / catch

try:
    litellm.completion(model="hf/org/model", messages=msgs)
except ValueError as e:
    if "No provider mapping found" in str(e):
        # permanent condition — switch model or self-host, do not retry
        raise RuntimeError("model not served by any HF inference provider") from e
    raise

Prevention

When it happens

Trigger: Requesting a model that has no Inference Providers enabled — gated/private models without provider deployment, freshly uploaded models not yet deployed, or datasets/spaces ids mistakenly used as model ids. The Hub returns 200 with metadata lacking inferenceProviderMapping.

Common situations: Using a niche or newly published model no provider hosts yet; model id typo that still resolves to some Hub object (e.g. wrong repo type); relying on a model whose provider deployments were removed after a provider partnership ended.

Related errors

AI-assisted analysis of BerriAI/litellm@6c2dcb801b (2026-08-15). Data as JSON: /api/errors/9a6d5b751e6d6d9d. Report an issue: GitHub.