microsoft/semantic-kernel · error · ServiceInitializationError

The MistralAI embedding model ID is required.

Error message

The MistralAI embedding model ID is required.

What it means

Raised as ServiceInitializationError right after settings validation succeeds, when `mistralai_settings.embedding_model_id` is still falsy. The connector requires a concrete embedding model id and will not guess one (unlike NVIDIA which has a default). It fires whether or not an api_key is present.

Source

Thrown at python/semantic_kernel/connectors/ai/mistral_ai/services/mistral_ai_text_embedding.py:69

            async_client : The Mistral AI client to use.
            env_file_path : The path to the environment file.
            env_file_encoding : The encoding of the environment file.

        Raises:
            ServiceInitializationError: If an error occurs during initialization.
        """
        try:
            mistralai_settings = MistralAISettings(
                api_key=api_key,
                embedding_model_id=ai_model_id,
                env_file_path=env_file_path,
                env_file_encoding=env_file_encoding,
            )
        except ValidationError as e:
            raise ServiceInitializationError(f"Failed to validate Mistral AI settings: {e}") from e

        if not mistralai_settings.embedding_model_id:
            raise ServiceInitializationError("The MistralAI embedding model ID is required.")

        if not async_client:
            async_client = Mistral(
                api_key=mistralai_settings.api_key.get_secret_value(),
            )
        super().__init__(
            service_id=service_id or mistralai_settings.embedding_model_id,
            ai_model_id=ai_model_id or mistralai_settings.embedding_model_id,
            async_client=async_client,
        )

    @override
    async def generate_embeddings(
        self,
        texts: list[str],
        settings: "PromptExecutionSettings | None" = None,
        **kwargs: Any,
    ) -> ndarray:

View on GitHub (pinned to c028a0c7dc)

Solutions

1. Pass ai_model_id explicitly: MistralAITextEmbedding(ai_model_id='mistral-embed').
2. Set MISTRALAI_EMBEDDING_MODEL_ID in env/.env (note the EMBEDDING_ segment).
3. Check for typos against the documented env var name.

Example fix

# before
svc = MistralAITextEmbedding(api_key=key)

# after
svc = MistralAITextEmbedding(api_key=key, ai_model_id='mistral-embed')

model_id = os.environ.get('MISTRALAI_EMBEDDING_MODEL_ID') or 'mistral-embed'
assert model_id, 'embedding model id required'
svc = MistralAITextEmbedding(ai_model_id=model_id, api_key=os.environ['MISTRALAI_API_KEY'])

def has_embedding_model_id(svc_cls, **kw) -> bool:
    mid = kw.get('ai_model_id') or os.environ.get('MISTRALAI_EMBEDDING_MODEL_ID')
    return bool(mid)

from semantic_kernel.exceptions import ServiceInitializationError
try:
    svc = MistralAITextEmbedding()
except ServiceInitializationError as e:
    if 'embedding model ID is required' in str(e):
        svc = MistralAITextEmbedding(ai_model_id='mistral-embed', api_key=os.environ['MISTRALAI_API_KEY'])
    else:
        raise

When it happens

Trigger: Constructing `MistralAITextEmbedding()` (or with only api_key) without supplying `ai_model_id` AND without setting MISTRALAI_EMBEDDING_MODEL_ID in env/.env. Validation passed (api_key ok) but the model id resolved to None.

Common situations: Developer copied the chat completion setup (which may default model id) expecting embeddings to do the same; MISTRALAI_EMBEDDING_MODEL_ID typo'd as MISTRALAI_MODEL_ID or MISTRALAI_CHAT_MODEL_ID.

Related errors

• Failed to validate Mistral AI settings: {e}
• Failed to create NVIDIA settings.
• The Amazon Bedrock Text Embedding Model ID is missing.
• Failed to validate Google AI settings: {e}
• The Google AI embedding model ID is required.

AI-assisted analysis of microsoft/semantic-kernel@c028a0c7dc (2026-08-13). Data as JSON: /api/errors/493f27f244fa3ad0. Report an issue: GitHub.