Unhelpful documentation to "LLM Provider not provided"

[otheus]

Issue

Version 0.86.1

Error message:

litellm.BadRequestError: LLM Provider NOT provided. Pass in the LLM provider you are trying to call. You passed model=local/qwen3-coder:30b
 Pass model as E.g. For 'Huggingface' inference endpoints pass in `completion(model='huggingface/starcoder',..)` Learn more: 
https://docs.litellm.ai/docs/providers

the error message makes no sense. The page at the given URL to "providers" makes no sense.
The responses [here|https://stackoverflow.com/a/79119200/3849157] are not helpful.
https://aider.chat/docs/llms/warnings.html isn't helpful here.

Example usage:

aider --model local/qwen3-coder:30b

Starts fine, but generates error on pretty much anything.

Models config:

cat $HOME/.aider.model.metadata.json
{ 
  "local/qwen3-coder:30b": { 
    "max_tokens": 16384,
    "max_output_tokens": 16384,
    "max_input_tokens": 65535,
    "input_cost_per_token": 0,
    "output_cost_per_token": 0,
    "supports_vision": "false",
    "litellm_provider": "local",
    "mode": "chat"
  }
}

It turns out that I was trying to use OPENAI_API_BASE instead of OLLAMA_API_BASE. Then, I was trying to give the base with /v1 and then / at the end, both of which caused it to croak. And I needed to replace local with ollama_chat, export the OLLAMA_API_BASE as directed here, but it isn't clear why. The first part is magical??

Related question: Wouldn't it make sense for the metadata here to provide a default api-base? eg

    ... 
    "default_api_base": "http://192.168.8.101:11434",
    ...

Version and model info

No response