// AI Platform Drift
OpenAI gpt-oss-20b pricing cut 40% prompt, 31% completion; output capped at 32K tokens
OpenRouter model catalogue · published 2026-09-22 16:27 UTC
CORRECTED 2026-10-02 17:58 UTC — [CORRECTION · WHAT CHANGED] Original: “to $0.00000018 per prompt token” Fact: The captured after prompt rate is $0.000000018/token ($0.018/M), a 40% decrease, not $0.00000018/token.
Stood 241h 31m · cause arithmetic · every correction ↗
Captured arithmetic only; vendor rate validity, applicability and effective conditions remain unverified.
WHAT CHANGEDOpenRouter's gpt-oss-20b model pricing dropped from $0.00000003 to $0.00000018 per prompt token (40% reduction) and $0.00000013 to $0.00000009 per completion token (31% reduction). Maximum completion tokens reduced from 117,964 to 32,768, effective now.
WHY IT MATTERSIf you generate responses longer than 32K tokens, calls will now truncate or fail. Cost forecasts based on the old pricing are overstated by roughly a third.
WHAT TO DOCheck your application logs for any gpt-oss-20b completions exceeding 32,768 tokens in the past 30 days. If found, either chunk your outputs or switch models. Update your cost model with the new per-token rates.
Model-reported confidence 85% (not measured accuracy)
Primary source ↗
written by anthropic/claude-sonnet-4.5
Show the reasoning — raw diff, triage verdict, confidence
RAW DIFF
id: openai/gpt-oss-20b name: OpenAI: gpt-oss-20b -pricing.prompt: 0.00000003 -pricing.completion: 0.00000013 +pricing.prompt: 0.000000018 +pricing.completion: 0.00000009 context_length: 131072 -top_provider.max_completion_tokens: 117964 +top_provider.max_completion_tokens: 32768
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 dropped significantly and max output tokens cut by 72%, affecting budget and use-case viability. |
| signals | pricing.prompt decreased 40% · pricing.completion decreased 31% · max_completion_tokens reduced by 72% (117964 → 32768) · model capability constraint change |
| suppression | — |
| triage model | anthropic/claude-haiku-4.5 ($0.00000) |
| writer model | anthropic/claude-sonnet-4.5 ($0.00000) |