How does max_tokens work on SuperCosmos?

max_tokens (or max_completion_tokens) sets the most tokens the model may write in one reply. On Max and Heavy that includes reasoning tokens, so a small value can be used up by thinking before the answer starts.

• Each model has a cap: GET /v1/models lists max_output_tokens (32,768 for Flash and Max, 49,152 for Heavy).
• If the reply hits the limit, finish_reason is length and the text is cut off.
• You pay only for tokens actually written, not for the limit you set.

Tips: leave generous room on Max and Heavy; on Flash a tight limit works well for short answers.

Good to know#

• New accounts get $5 of free credit; top up from $1 to $800 per purchase in Console › Billing › Credits. Billing
• Retry only 429 and 5xx errors, with backoff; honour Retry-After when it's sent. Which errors to retry
• Every response has an X-Request-Id; log it so a request can be found later. Why log it

Still need help?#

• General support: support@clatterlabs.com or team@clatterlabs.xyz
• Infrastructure questions, or permanent deletion of your data (your whole account or one service): infrastructure@clatterlabs.xyz

We usually reply within the first hour, and always within 24 hours. A full support team is with you.