// AI Platform Drift
DeepSeek Pro Latest: completion price cut 60%, output limit reduced 58%
OpenRouter model catalogue · published 2026-09-24 13:44 UTC
WHAT CHANGEDCompletion pricing dropped from $0.0000029 to $0.0000011682 per token (60% reduction). Prompt pricing decreased slightly from $0.00000039 to $0.0000003894. Maximum completion tokens reduced from 943,718 to 393,216 tokens (58% reduction). Changes are live as of 24 September 2026.
WHY IT MATTERSIf you generate responses longer than 393K tokens, calls will now fail or truncate where they previously succeeded. The cost reduction is substantial for high-volume completion workloads, but the output limit is now less than half what it was.
WHAT TO DOCheck your application logs for any DeepSeek Pro responses exceeding 393,216 tokens in the past 30 days. If found, either chunk your requests or switch to a model with higher limits. Reforecast API costs if DeepSeek Pro represents >10% of your completion token 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-pro-latest name: DeepSeek: DeepSeek Pro Latest -pricing.prompt: 0.00000039 -pricing.completion: 0.0000029 +pricing.prompt: 0.0000003894 +pricing.completion: 0.0000011682 context_length: 1048576 -top_provider.max_completion_tokens: 943718 +top_provider.max_completion_tokens: 393216
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 and token limit reduction directly affect engineering budgets and model capacity planning. |
| signals | prompt pricing decreased · completion pricing decreased · max_completion_tokens reduced by 58% |
| suppression | — |
| triage model | anthropic/claude-haiku-4.5 ($0.00000) |
| writer model | anthropic/claude-sonnet-4.5 ($0.00000) |