OpenAI
gpt-oss-120b
openai/gpt-oss-120b
输入
$0.15
每 1M token
输出
$0.60
每 1M token
缓存输入
$0.075
每 1M token
gpt-oss-120b is an open-weight, 117B-parameter Mixture-of-Experts (MoE) language model from OpenAI designed for high-reasoning, agentic, and general-purpose production use cases. It activates 5.1B parameters per forward pass and is optimized...
以下描述来自 OpenRouter，原文为英文。
- 发布时间
- 上下文窗口
- 131K token
- 最大输出
- 66K token
- 输入
- 文本
- 输出
- 文本
- 推理能力
- 支持
价格变动历史
自开始追踪以来未发生价格变动（追踪始于 .
OpenRouter 上的服务商
当前有 25 个接口提供此模型。正常运行率为 OpenRouter 上的请求成功率；价格单位为每 1M token。
| 服务商 | 输入 | 输出 | 上下文窗口 | 30 分钟正常率 | 24 小时正常率 |
|---|---|---|---|---|---|
| Near AIfp4 | $0.03 | $0.15 | 131K | 100% | 100% |
| Veniceint4 | $0.03 | $0.15 | 128K | 96.7% | 99.72% |
| CoreWeavefp4 | $0.03 | $0.17 | 131K | 100% | 100% |
| DeepInfrabf16 | $0.037 | $0.17 | 131K | 100% | 98.8% |
| DekaLLMbf16 | $0.03 | $0.18 | 131K | 100% | 99.93% |
| AkashMLbf16 | $0.03 | $0.187 | 131K | 100% | 100% |
| Crusoebf16 | $0.05 | $0.25 | 131K | 100% | 100% |
| Novitafp4 | $0.05 | $0.25 | 131K | 100% | 99.72% |
| Mancer 2fp8 | $0.05 | $0.25 | 131K | 100% | 99.58% |
| Googleglobal | $0.09 | $0.36 | 131K | 0.0% | 2.6% |
| Sail Researchfp4 | $0.06 | $0.40 | 131K | 100% | 100% |
| DigitalOcean | $0.06 | $0.42 | 128K | 100% | 100% |
| BaseTenfp4 · us | $0.10 | $0.50 | 128K | 100% | 99.01% |
| BaseTenfp4 | $0.10 | $0.50 | 128K | 100% | 98.9% |
| Amazon Bedrock | $0.15 | $0.60 | 131K | 100% | 100% |
| Amazon Bedrockeu-west-1 | $0.15 | $0.60 | 131K | 100% | 100% |
| Groq | $0.15 | $0.60 | 131K | 100% | 100% |
| Nebiusfp4 | $0.15 | $0.60 | 131K | 100% | 100% |
| Phala | $0.15 | $0.60 | 131K | 100% | 99.44% |
| Together | $0.15 | $0.60 | 131K | 100% | 98.7% |
| SiliconFlowfp8 | $0.15 | $0.60 | 131K | 66.7% | 96.4% |
| Parasailfp4 | $0.10 | $0.75 | 131K | 100% | 100% |
| Mara | $0.15 | $0.75 | 131K | 100% | 97.3% |
| Cerebrasfp16 | $0.35 | $0.75 | 131K | 100% | 100% |
| SambaNova | $0.14 | $0.95 | 131K | 100% | 99.92% |