// AI Platform Drift

DeepSeek V4.1 Flash pricing dropped 75%, output limit increased 140%

OpenRouter model catalogue · published 2026-09-25 16:44 UTC
WORTH KNOWING BUDGET PRICINGLIMITSMODEL
WHAT CHANGEDOpenRouter's deepseek-v4.1-flash model pricing decreased from $0.0000003 to $0.000000075 per prompt token and $0.0000012 to $0.0000003 per completion token. Maximum completion tokens increased from 393,216 to 943,718.
WHY IT MATTERSIf you are using this model, your inference costs just dropped by three-quarters and you can now generate outputs 2.4x longer without hitting the provider limit. Existing cost projections and output-length assumptions are now wrong.
WHAT TO DORerun your cost model with the new rates and update any code that caps output length below 943,718 tokens if longer responses would be useful.
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-v4.1-flash
 name: DeepSeek: DeepSeek V4.1 Flash
-pricing.prompt: 0.0000003
-pricing.completion: 0.0000012
+pricing.prompt: 0.000000075
+pricing.completion: 0.0000003
 context_length: 1048576
-top_provider.max_completion_tokens: 393216
+top_provider.max_completion_tokens: 943718
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 reasoningSignificant price reductions and token limit increase affect cost and capability planning.
signalsprompt pricing decreased 75% (0.0000003 → 0.000000075) · completion pricing decreased 75% (0.0000012 → 0.0000003) · max_completion_tokens increased 140% (393216 → 943718)
suppression—
triage modelanthropic/claude-haiku-4.5 ($0.00000)
writer modelanthropic/claude-sonnet-4.5 ($0.00000)