openai/gpt-6.1-sol,
anthropic/claude-opus-5-5,
or gemini/gemini-3.8-flash.
You can also try a custom model identifier such as
fireworks_ai/accounts/my-account/models/my-custom-model.
The result comes from LiteLLM’s get_llm_provider(). A successful lookup does
not verify that the model exists or that you have access to it. Provider adapters
can further transform the parsed model name when making an inference request.
The chatgpt and github_copilot providers are unavailable in this lookup because
their parsers require authentication.