Skip to main content
Code: missing_max_tokens_no_model_default HTTP status: 400 Type: invalid_request

​
When it fires

POST /v1/chat/completions was called without max_tokens, AND the model row has no default_max_output_tokens configured for that case. Most models advertise a sensible default (aurous_metadata.default_max_output_tokens); this code surfaces only on models that intentionally require explicit caps.

​
How to recover

Pass max_tokens explicitly on the request. Pick a value at or below the model’s max_output_tokens_hard_cap. Re-read the model row from GET /v1/models to see the hard cap.