// AI Platform Drift
DeepSeek V4 Flash completion token limit cut from 384k to 131k
OpenRouter model catalogue · published 2026-09-27 11:33 UTC
CORRECTED 2026-10-02 17:58 UTC — [CORRECTION · WHAT CHANGED] Original: “Completion pricing dropped 65.9% ($384,000/M to $131,072/M tokens)” Fact: Completion pricing falls 0.2976%, from $0.09408/M to $0.0938/M. Output-token capacity falls 65.9%.
[CORRECTION · WHY IT MATTERS] Original: “The pricing drop is substantial” Fact: The captured price reduction is approximately 0.3%; the output-capacity reduction is 65.9%.
Stood 126h 24m · cause extraction error · every correction ↗
[ARITHMETIC NOTE · triage.signals] Original: “pricing.prompt decreased ~0.5%” Fact: Prompt pricing decreases 0.2976% (approximately 0.3%).
[ARITHMETIC NOTE · triage.signals] Original: “pricing.completion decreased ~0.5%” Fact: Completion pricing decreases 0.2976% (approximately 0.3%).
Captured arithmetic only; vendor rate validity, applicability and effective conditions remain unverified.
WHAT CHANGEDOpenRouter's DeepSeek V4 Flash (deepseek/deepseek-v4-flash) reduced max_completion_tokens from 384,000 to 131,072 tokens. Completion pricing dropped 65.9% ($384,000/M to $131,072/M tokens) and prompt pricing dropped 0.3% ($0.047/M to $0.0469/M tokens).
WHY IT MATTERSApplications generating outputs longer than 131,072 tokens will now fail or truncate. The pricing drop is substantial but does not offset the hard limit reduction if your workload depends on the previous 384k completion capacity.
WHAT TO DOGrep your codebase for 'deepseek-v4-flash' or 'deepseek/deepseek-v4-flash' and check max_tokens or equivalent parameters in calls. If any exceed 131,072, either chunk the generation into multiple requests or switch to a model with higher completion limits.
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: 0.00000004704 -pricing.completion: 0.00000009408 +pricing.prompt: 0.0000000469 +pricing.completion: 0.0000000938 context_length: 1048576 -top_provider.max_completion_tokens: 384000 +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 verdict | MATERIAL at 0.95, needed 0.5 |
|---|---|
| triage reasoning | Pricing change and significant reduction in max output tokens affect budgeting and capability planning. |
| signals | pricing.prompt decreased ~0.5% · pricing.completion decreased ~0.5% · max_completion_tokens reduced by 66% (384000 → 131072) |
| suppression | — |
| triage model | anthropic/claude-haiku-4.5 ($0.00000) |
| writer model | anthropic/claude-sonnet-4.5 ($0.00000) |