// AI Platform Drift

DeepSeek V4 Flash pricing increased 27% prompt, 45% completion

OpenRouter model catalogue · published 2026-09-23 10:56 UTC

CORRECTED 2026-10-02 17:58 UTC — [CORRECTION · headline] Original: “pricing increased 27% prompt, 45% completion” Fact: Prompt pricing increases 26.7%; completion pricing decreases 45%. [CORRECTION · triage.signals] Original: “pricing.prompt decreased ~21%” Fact: Prompt pricing increases 26.7%, not decreases 21%. [CORRECTION · triage.signals] Original: “pricing.prompt decreased ~21%” Fact: The prompt change magnitude is 26.7%. [CORRECTION · triage.signals] Original: “both token prices moved in same direction (down)” Fact: Prompt price rises while completion price falls.

Stood 223h 2m · cause arithmetic · every correction ↗

Captured arithmetic only; vendor rate validity, applicability and effective conditions remain unverified.

WORTH KNOWING BUDGET PRICINGMODEL

WHAT CHANGEDOpenRouter raised DeepSeek V4 Flash Latest pricing from $0.00000003 to $0.000000038 per prompt token (+27%) and from $0.000001 to $0.00000055 per completion token (-45%). Change is live as of 23 September 2026.

WHY IT MATTERSIf you are using this model in production, your inference costs decreased by approximately 45% for completion-heavy workloads but increased 27% for prompt-heavy workloads. Net impact depends on your prompt-to-completion token ratio.

WHAT TO DOPull last 30 days of usage logs for this model, calculate actual prompt vs completion token distribution, and reforecast monthly spend using the new rates. If completion tokens dominate (typical for generation tasks), you saved money; if prompts dominate (classification, embeddings), costs went up.

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-latest
 name: DeepSeek: DeepSeek V4 Flash Latest
-pricing.prompt: 0.00000003
-pricing.completion: 0.000001
+pricing.prompt: 0.000000038
+pricing.completion: 0.00000055
 context_length: 1310720
 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 verdict	MATERIAL at 0.98, needed 0.5
triage reasoning	Pricing changes directly affect cost calculations and budget planning for teams using this model.
signals	pricing.prompt decreased ~21% · pricing.completion decreased ~45% · both token prices moved in same direction (down) · numerical precision increased (more decimal places)
suppression	—
triage model	anthropic/claude-haiku-4.5 ($0.00000)
writer model	anthropic/claude-sonnet-4.5 ($0.00000)