// AI Platform Drift

DeepSeek V4 Flash pricing reduced 25.6% on both input and output

OpenRouter model catalogue · published 2026-09-25 14:06 UTC
WORTH KNOWING BUDGET PRICINGMODEL
WHAT CHANGEDOpenRouter lowered DeepSeek V4 Flash 0423 pricing from $0.06552 to $0.04872 per million prompt tokens and from $0.13104 to $0.09744 per million completion tokens, effective now.
WHY IT MATTERSIf you are using this model, your inference costs dropped by roughly a quarter with no action required. If you chose a different model because DeepSeek was too expensive, the economics may now favor switching.
WHAT TO DOPull last month's token usage for this model from your logs or OpenRouter dashboard, multiply by the new rates, and update your cost forecast if this model represents material spend.
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-flash
 name: DeepSeek: DeepSeek V4 Flash 0423
-pricing.prompt: 0.00000006552
-pricing.completion: 0.00000013104
+pricing.prompt: 0.00000004872
+pricing.completion: 0.00000009744
 context_length: 1048576
 top_provider.max_completion_tokens: 384000
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 reasoningPrice reduction on active model affects cost planning and budget forecasts for teams using this model.
signalspricing.prompt decreased 25.6% · pricing.completion decreased 25.6% · both token prices lowered symmetrically
suppression—
triage modelanthropic/claude-haiku-4.5 ($0.00000)
writer modelanthropic/claude-sonnet-4.5 ($0.00000)