OpenRouter max completion tokens reduced from 943,718 to 131,072

OpenRouter model catalogue · published 2026-09-23 03:42 UTC

ACTION NEEDED VERIFY_ONLY LIMITSAPIMODEL

WHAT CHANGEDThe top_provider.max_completion_tokens field dropped from 943,718 to 131,072 tokens—an 86% reduction. Context length remains 1,310,720 tokens. Pricing unchanged at $0.000000075 per prompt token and $0.00000025 per completion token.

WHY IT MATTERSAny request attempting to generate more than 131,072 tokens will now fail or be truncated. If you rely on long-form generation (summarizing large documents, generating extensive code, multi-turn conversations with large outputs), existing calls may break without warning.

WHAT TO DOGrep your codebase for max_tokens or max_completion_tokens parameters sent to OpenRouter. If any exceed 131,072, reduce them or split the work into multiple requests. Test one long-output workflow end-to-end this week.

CONFIDENCE 95% Primary source ↗ written by anthropic/claude-sonnet-4.5

Show the reasoning — raw diff, triage verdict, confidence

RAW DIFF

 pricing.prompt: 0.000000075
 pricing.completion: 0.00000025
 context_length: 1310720
-top_provider.max_completion_tokens: 943718
+top_provider.max_completion_tokens: 131072

HOW IT WAS JUDGED

triage verdict	MATERIAL at 0.95, needed 0.5
triage reasoning	Model's maximum completion token output reduced by 86%, directly impacts application design and token budgeting.
signals	max_completion_tokens decreased significantly · 943718 → 131072 is ~7x reduction · affects model capability / output limits
suppression	—
triage model	anthropic/claude-haiku-4.5 ($0.00000)
writer model	anthropic/claude-sonnet-4.5 ($0.00000)