// AI Platform Drift

DeepSeek V4 Flash 0731 completion tokens up 300%, prompt tokens down 44%

OpenRouter model catalogue · published 2026-09-30 21:59 UTC
ACTION NEEDED BUDGET PRICINGMODEL
WHAT CHANGEDOpenRouter changed pricing for deepseek/deepseek-v4-flash-0731 effective now. Prompt tokens dropped from $0.0180/M to $0.0100/M (-44%). Completion tokens rose from $0.3200/M to $1.2800/M (+300%).
WHY IT MATTERSTotal cost per request increases 282% to 298% depending on output length, because completion tokens cost four times more while prompt savings are small. A workload generating ten completion tokens per prompt token now costs 298% more.
WHAT TO DORerun cost projections using the new rates and update budgets. If this model is in production, calculate actual spend impact from last 30 days of traffic and escalate the forecast to whoever owns the API budget.
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-0731
 name: DeepSeek: DeepSeek V4 Flash 0731
-pricing.prompt: 0.000000018
-pricing.completion: 0.00000032
+pricing.prompt: 0.00000001
+pricing.completion: 0.00000128
 context_length: 1048576
 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 0.95, needed 0.5
triage reasoningPricing changes directly affect cost calculations and budget planning for production workloads.
signalspricing.prompt decreased 44% (0.000000018 → 0.00000001) · pricing.completion increased 300% (0.00000032 → 0.00000128) · asymmetric price movement across token types
suppression—
triage modelanthropic/claude-haiku-4.5 ($0.00000)
writer modelanthropic/claude-sonnet-4.5 ($0.00000)