Operations
Model requests for 2026-11-12 (UTC). Actual provider tokens are used when available. Unknown or interrupted requests keep their input estimate and output reservation.
Usage limits
Per operation: 300 HTTP attempts · 2,000,000 input tokens · 200,000 output tokens.
Daily token ceiling: disabled. Retries of the same operation share its budget.
Preflight input allowance uses one token per serialized character plus 4,096 per image. This is a conservative allowance, not an exact tokenizer. Dollar amounts require configured input and output prices.
Today’s model usage
No recorded model requests today.