Fosowl/agenticSeek · error · Exception

LiteLLM is not available for local use. Change config.ini

Error message

LiteLLM is not available for local use. Change config.ini

What it means

litellm_fn explicitly rejects local configurations: if the provider was constructed with is_local=True (a local model chosen in config.ini), it raises a plain Exception saying LiteLLM routing is only for cloud providers. LiteLLM completion here always targets hosted APIs via model prefixes, so local backends are unsupported by design.

Source

Thrown at sources/llm_provider.py:522

            raise NetworkError("Network error occurred. Check your internet connection.") from e
        except APIError as e:
            raise APIError(f"API error occurred: {str(e)}") from e
        return None

    def litellm_fn(self, history, verbose=False):
        """
        Use LiteLLM AI gateway for completion.
        Routes to 100+ providers (OpenAI, Anthropic, Azure, Bedrock,
        Vertex AI, Groq, Together, Ollama, etc.) based on model prefix.
        See https://docs.litellm.ai/docs/providers
        """
        try:
            import litellm
        except ImportError as e:
            raise ImportError("litellm is not installed. Install with: pip install litellm") from e

        if self.is_local:
            raise Exception("LiteLLM is not available for local use. Change config.ini")

        api_key = os.getenv("LITELLM_API_KEY", None)

        try:
            call_kwargs = {
                "model": self.model,
                "messages": history,
                "drop_params": True,
            }
            if api_key:
                call_kwargs["api_key"] = api_key
            response = litellm.completion(**call_kwargs)
            if response is None:
                raise Exception("LiteLLM response is empty.")
            thought = response.choices[0].message.content
            if verbose:
                print(thought)
            return thought

View on GitHub (pinned to ae57a23577)

Solutions

1. Edit config.ini and switch the provider away from local mode (set is_local/local flag off and pick a cloud model)
2. Use the dedicated local-model code path in LLMProvider instead of litellm_fn
3. Instantiate a separate LLMProvider configured for a cloud provider when you need LiteLLM routing

Example fix

// before (config.ini)
[PROVIDER]
provider = litellm
local = true
// after (config.ini)
[PROVIDER]
provider = litellm
local = false
model = openai/gpt-4o-mini

if provider.is_local:
    raise SystemExit('litellm_fn requires a cloud provider: set local = false in config.ini')

try:
    thought = provider.litellm_fn(history)
except Exception as e:
    if 'not available for local use' in str(e):
        logger.error('Fix config.ini: LiteLLM cannot serve local models here')
    raise

When it happens

Trigger: Calling LLMProvider.litellm_fn(history) when the provider instance was built from a config.ini entry selecting a local model (is_local=True).

Common situations: Developer flips the provider to litellm in config.ini but leaves `local = true` / local model settings; reusing one provider object across local and cloud pipelines; misunderstanding that litellm can proxy Ollama-style local endpoints in this wrapper.

Related errors

• Model not set
• Prompt file not found at path: {file_path}
• Permission denied to read prompt file at path: {file_path}
• Unknown provider: {provider_name}
• API key {api_key_var} not found in .env file. Please add it

AI-assisted analysis of Fosowl/agenticSeek@ae57a23577 (2026-08-30). Data as JSON: /api/errors/213edf97e86bef65. Report an issue: GitHub.