io.github.daedalus/mcp-llm-gateway

List all available models from the configured providers. Fetches models from the configured providers, with caching. Can filter by provider ID. Args: provider: Optional provider ID to filter models. Returns: List of model objects with id, name, and metadata. Example: >>> list_models() [{"id": "gpt-4", "object": "model", "created": 1234567890, "owned_by": "openai", "provider_id": "openai"}] >>> list_models(provider="anthropic") [{"id": "claude-3", "object": "model", "provider_id": "anthropic"}]

Send a completion request to the downstream LLM provider. Proxies the request to the configured OpenAI-compatible downstream endpoint. Args: prompt: The input prompt for the model. model: Optional model ID. Uses provider default if not specified. provider: Optional provider ID. Uses first enabled provider if not specified. max_tokens: Optional maximum tokens to generate. temperature: Optional sampling temperature (0.0 to 2.0). Returns: Completion response from the downstream provider. Example: >>> complete("Hello, world!", model="gpt-4", max_tokens=100) {"id": "...", "object": "chat.completion", ...} >>> complete("Hello", provider="anthropic", model="claude-3") {"id": "...", "object": "chat.completion", ...}