// AI Platform Drift

DeepSeek Flash completion price cut 3%, output limit reduced to 131K tokens

OpenRouter model catalogue · published 2026-10-16 21:11 UTC
ACTION NEEDED VERIFY_ONLY LIMITSPRICINGMODEL
WHAT CHANGEDOpenRouter's ~deepseek/deepseek-flash-latest model changed on 2026-10-16: captured completion base rate changed from $0.6633 to $0.6435 per 1M tokens, and max_completion_tokens dropped from 943,718 to 131,072 tokens.
WHY IT MATTERSRequests generating more than 131,072 completion tokens will now fail or truncate. Using the captured base rates for ~deepseek/deepseek-flash-latest, 100M completion tokens per day cost $66.33 before and $64.35 after, excluding all other charges.
WHAT TO DOGrep your codebase for max_tokens or max_completion_tokens parameters set above 131,072 when calling this model, and reduce them to stay within the new limit.
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: ~deepseek/deepseek-flash-latest
 name: DeepSeek: DeepSeek Flash Latest
 pricing.prompt: 0.000000036
-pricing.completion: 0.0000006633
+pricing.completion: 0.0000006435
 context_length: 1048576
-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 0.95, needed 0.5
triage reasoningPricing decreased slightly; output token limit cut dramatically, affecting max request size.
signalscompletion token price decreased · max_completion_tokens limit reduced by 86% (943718 → 131072) · pricing field changed · output limit changed
suppression—
triage modelanthropic/claude-haiku-4.5 ($0.00000)
writer modelanthropic/claude-sonnet-4.5 ($0.00000)