gpt-oss-20b pricing and routing profile
gpt-oss-20b from OpenAI lists at $0.018 per million input tokens and $0.09 per million output tokens, with a 131,072-token context window. Prices from OpenRouter's catalogue on 2026-09-28.
Tool callingStructured outputReasoning controlLog probabilities
Cost per 1,000 requests
| Request shape | List cost |
|---|---|
| Classification or routing (600 tokens in, 10 out) | $0.0117 |
| Chat reply (2,000 in, 400 out) | $0.072 |
| Retrieval-augmented answer (8,000 in, 500 out) | $0.189 |
Before prompt caching. Try your own shape in the cost calculator.
Can Gimbix route to it first?
Yes. 6 of 12 tracked hosts return token log probabilities for gpt-oss-20b, which is the confidence signal Gimbix uses to let a cheaper model answer first.
Measured by Gimbix
On 400 real BANKING77 customer messages, gpt-oss-20b answered 313 correctly (95% interval 73.9%–82.0%), with 1 invalid answers, at $0.014 per 1,000 requests and a median of 1,355 ms. See all 22 models.
Who serves gpt-oss-20b
| Host | Input / output per 1M | Log probabilities | Uptime, last 30 min |
|---|---|---|---|
| Darkbloom | $0.018 / $0.09 | yes | 99.8% |
| AkashML | $0.02 / $0.10 | yes | 99.8% |
| DekaLLM | $0.029 / $0.14 | yes | 98.8% |
| CoreWeave | $0.03 / $0.13 | yes | 100.0% |
| DeepInfra | $0.03 / $0.14 | no | 99.6% |
| Parasail | $0.03 / $0.15 | yes | 99.9% |
| Novita | $0.04 / $0.15 | yes | 99.9% |
| SiliconFlow | $0.04 / $0.18 | no | 97.1% |
| Amazon Bedrock | $0.07 / $0.15 | no | — |
| Amazon Bedrock | $0.07 / $0.15 | no | 99.8% |
| $0.07 / $0.25 | no | 68.3% | |
| Groq | $0.075 / $0.30 | no | 97.8% |
A host that lists log probabilities may still omit them; Gimbix checks each one before relying on it.
Compare gpt-oss-20b
- gpt-oss-20b vs Claude Opus 5.5
- gpt-oss-20b vs Claude Sonnet 5
- gpt-oss-20b vs Claude Fable 5.1
- gpt-oss-20b vs Claude Opus 5
- gpt-oss-20b vs GPT-6 Sol
- gpt-oss-20b vs GPT-6 Astra
- gpt-oss-20b vs GPT-5.6 Sol
- gpt-oss-20b vs GPT-5.6 Terra
- gpt-oss-20b vs GPT-5.5
- gpt-oss-20b vs GPT-5.4
- gpt-oss-20b vs Gemini 3.1 Pro Preview
- gpt-oss-20b vs Gemini 3.8 Flash
- gpt-oss-20b vs Grok 4.7
- gpt-oss-20b vs Qwen3.8 Max (0902)
- gpt-oss-20b vs Kimi K3
- gpt-oss-20b vs GLM 5.3
- gpt-oss-20b vs DeepSeek V4 Pro 0423
- gpt-oss-20b vs Mistral Large 3 2512
- gpt-oss-20b vs Nova Premier 1.0
- gpt-oss-20b vs Command A
- gpt-oss-20b vs GPT-6 Luna
- gpt-oss-20b vs GPT-5.6 Luna
- gpt-oss-20b vs GPT-5.4 Mini
- gpt-oss-20b vs GPT-5.4 Nano
- gpt-oss-20b vs GPT-4.1 Mini
- gpt-oss-20b vs gpt-oss-120b
- gpt-oss-20b vs Claude Haiku 4.5
- gpt-oss-20b vs Gemini 3.5 Flash Lite
- gpt-oss-20b vs Gemma 4 31B
- gpt-oss-20b vs DeepSeek V4.1 Flash
- gpt-oss-20b vs DeepSeek V4 Flash 0731
- gpt-oss-20b vs GLM 5.3 Flash
- gpt-oss-20b vs Hy4 preview
- gpt-oss-20b vs MiMo-V2.6-Flash
- gpt-oss-20b vs Nemotron 3.5 Lightning
- gpt-oss-20b vs Nemotron 3 Ultra
- gpt-oss-20b vs Qwen3.8 Flash
- gpt-oss-20b vs Qwen3.7 Flash
- gpt-oss-20b vs Qwen3.8 27B
- gpt-oss-20b vs MiniMax M3
- gpt-oss-20b vs Kimi K2.6
- gpt-oss-20b vs Mistral Small 4
- gpt-oss-20b vs Ministral 3 8B 2512
- gpt-oss-20b vs Llama 4 Maverick
- gpt-oss-20b vs Llama 3.3 70B Instruct
- gpt-oss-20b vs Nova Lite 1.0
- gpt-oss-20b vs Nova Micro 1.0
- gpt-oss-20b vs Command R7B (12-2024)
- gpt-oss-20b vs Phi 4
Questions
How much does gpt-oss-20b cost?
gpt-oss-20b lists at $0.018 per million input tokens and $0.09 per million output tokens on OpenRouter as of 2026-09-28. A typical chat reply of 2,000 input and 400 output tokens costs about $0.072 per 1,000 requests.
What is the context window of gpt-oss-20b?
131,072 tokens.
Can a cheaper model replace gpt-oss-20b?
Sometimes. Gimbix proves it on your own traffic before routing: the cheaper model must match gpt-oss-20b on one half of your requests and make at most three answers worse on the other.