BerriAI/litellm · error · ValueError

api_base is required for SambaNova embeddings

Error message

api_base is required for SambaNova embeddings

What it means

SambaNova embeddings have no default base URL baked into LiteLLM, so get_complete_url raises ValueError('api_base is required for SambaNova embeddings') when no api_base was resolved from the call, litellm params, or environment before the request is built.

Source

Thrown at litellm/llms/sambanova/embedding/transformation.py:34

from ..common_utils import SambaNovaError


class SambaNovaEmbeddingConfig(BaseEmbeddingConfig):
    def __init__(self) -> None:
        pass

    def get_complete_url(
        self,
        api_base: str | None,
        api_key: str | None,
        model: str,
        optional_params: dict,
        litellm_params: dict,
        stream: bool | None = None,
    ) -> str:
        if api_base is None:
            raise ValueError("api_base is required for SambaNova embeddings")
        # Remove trailing slashes and ensure clean base URL
        api_base = api_base.rstrip("/")
        if not api_base.endswith("/embeddings"):
            api_base = f"{api_base}/embeddings"
        return api_base

    def validate_environment(
        self,
        headers: dict,
        model: str,
        messages: list[AllMessageValues],
        optional_params: dict,
        litellm_params: dict,
        api_key: str | None = None,
        api_base: str | None = None,
    ) -> dict:
        if api_key is None:
            api_key = get_secret_str("SAMBANOVA_API_KEY")

View on GitHub (pinned to 77b7c6c40c)

Solutions

1. Pass api_base explicitly: litellm.embedding(model='sambanova/E5-Embeddings-300M', input=[...], api_base='https://api.sambanova.ai/v1').
2. Or set the environment variable SAMBANOVA_API_BASE (and SAMBANOVA_API_KEY) once per environment.
3. Or set litellm.api_base globally if all calls in the process target the same SambaNova deployment.
4. For self-hosted stacks, point api_base at your own gateway; LiteLLM appends /embeddings when missing.

Example fix

# before
litellm.embedding(model='sambanova/E5-Embeddings-300M', input=['hi'])
# after
litellm.embedding(
    model='sambanova/E5-Embeddings-300M',
    input=['hi'],
    api_base='https://api.sambanova.ai/v1',
)

import os

API_BASE = os.environ.get('SAMBANOVA_API_BASE') or 'https://api.sambanova.ai/v1'
assert API_BASE, 'SambaNova embeddings require an api_base'
out = litellm.embedding(model='sambanova/E5-Embeddings-300M', input=texts, api_base=API_BASE)

When it happens

Trigger: Calling litellm.embedding(model='sambanova/...', input=[...]) without api_base in the request, without litellm.api_base set, and without the SAMBANOVA_API_BASE environment variable.

Common situations: Assuming sambanova/ routes to a public URL automatically (only chat models have a default); migrating configs where api_base was set per-team via a shared litellm module the new code does not import; self-hosted SambaNova Enterprise deployments where the URL is site-specific.

Related errors

• api_base is required for PydanticAIProviderConfig
• api_base is required for Pydantic AI agents
• api_base is required for Pydantic AI agents
• api_key and api_base are required for Arize Phoenix prompt i
• api_base is required

AI-assisted analysis of BerriAI/litellm@77b7c6c40c (2026-08-18). Data as JSON: /api/errors/7e50966853004d12. Report an issue: GitHub.