// AI Platform Drift
DeepSeek V4 Flash Vision max completion tokens cut from 943K to 262K
OpenRouter model catalogue · published 2026-09-27 04:06 UTC
WHAT CHANGEDOpenRouter reduced max_completion_tokens for deepseek/deepseek-v4-flash-vision-exp from 943,718 to 262,144 tokens (72% reduction). Prompt pricing decreased from $0.00000022 to $0.0000002156 per token; completion pricing decreased from $0.00000066 to $0.0000006468 per token.
WHY IT MATTERSApplications requesting or expecting output longer than 262K tokens will now be truncated or rejected. The pricing changes are negligible (2% decrease), but the token limit reduction breaks any workflow that relied on the previous 943K ceiling for long-form generation.
WHAT TO DOGrep your codebase for 'deepseek-v4-flash-vision-exp' or 'deepseek/deepseek-v4-flash-vision-exp' and check max_tokens parameters in requests. If any exceed 262,144, either chunk the output into multiple calls or switch to a model with a higher completion limit.
Show the reasoning — raw diff, triage verdict, confidence
RAW DIFF
id: deepseek/deepseek-v4-flash-vision-exp name: DeepSeek: DeepSeek V4 Flash Vision Exp -pricing.prompt: 0.00000022 -pricing.completion: 0.00000066 +pricing.prompt: 0.0000002156 +pricing.completion: 0.0000006468 context_length: 1048576 -top_provider.max_completion_tokens: 943718 +top_provider.max_completion_tokens: 262144
HOW IT WAS JUDGED
| triage verdict | MATERIAL at 0.95, needed 0.5 |
|---|---|
| triage reasoning | Pricing dropped and output token limit cut significantly; teams budgeting or designing around this model need to know. |
| signals | pricing.prompt decreased · pricing.completion decreased · max_completion_tokens reduced by 72% |
| suppression | — |
| triage model | anthropic/claude-haiku-4.5 ($0.00000) |
| writer model | anthropic/claude-sonnet-4.5 ($0.00000) |