// AI Platform Drift

Z.ai GLM Latest max completion tokens cut from 943,718 to 131,072

OpenRouter model catalogue · published 2026-09-22 16:28 UTC
ACTION NEEDED VERIFY_ONLY LIMITSMODELAPI
WHAT CHANGEDThe top_provider.max_completion_tokens for model ~z-ai/glm-latest dropped from 943,718 to 131,072 tokens (86% reduction). Pricing changed negligibly: prompt tokens from $0.0000006545 to $0.0000006538, completion tokens from $0.000002057 to $0.0000020548 per token.
WHY IT MATTERSAny code requesting completion lengths above 131,072 tokens will now fail or be truncated. If you generate long-form content, transcripts, or batch outputs with this model, those requests will break.
WHAT TO DOGrep your codebase for max_tokens or max_completion_tokens parameters sent to ~z-ai/glm-latest. If any exceed 131,072, either reduce them or switch to a model with higher limits before your next deployment.
Model-reported confidence 95% (not measured accuracy) Primary source ↗ written by anthropic/claude-sonnet-4.5
Show the reasoning — raw diff, triage verdict, confidence
RAW DIFF
 id: ~z-ai/glm-latest
 name: Z.ai: GLM Latest
-pricing.prompt: 0.0000006545
-pricing.completion: 0.000002057
+pricing.prompt: 0.0000006538
+pricing.completion: 0.0000020548
 context_length: 1310720
-top_provider.max_completion_tokens: 943718
+top_provider.max_completion_tokens: 131072
HOW IT WAS JUDGED
Triage signals below are model-reported assertions, preserved verbatim; they are not independent arithmetic verification.
triage verdictMATERIAL at 0.98, needed 0.5
triage reasoningPricing dropped and max output tokens cut dramatically. Teams budgeting or relying on long completions need to know.
signalspricing.prompt decreased · pricing.completion decreased · max_completion_tokens reduced by 86% (943718 → 131072)
suppression—
triage modelanthropic/claude-haiku-4.5 ($0.00000)
writer modelanthropic/claude-sonnet-4.5 ($0.00000)