GPT-OSS 20B
Freeopenai/gpt-oss-20bText · 131K context · Apr 2025 · 24 upstreamsupstream上游
GPT-OSS 20B is OpenAI's open-weight reasoning model for agentic tasks and low-latency edge use, under Apache-2.0. It is a mixture-of-experts model with 20.9B total and 3.6B active parameters and a 131,072-token context. It runs within 16 GB of memory with MXFP4 weights.
- contextngữ cảnh上下文
- 131K
- inputđầu vào输入 / 1M
- $0
- outputđầu ra输出 / 1M
- $0
- routestuyến路由
- 24
- speed, 24htốc độ, 24h速度,24h
- No data yetChưa có dữ liệu暂无数据
Upstreams
24- 1PollinationsPlatformFree
- 2AIHubMixYour keynot addedBilled to your key
- 3Hugging FaceYour keynot addedBilled to your key
- 4DeepInfraYour keynot addedBilled to your key
- 5ClineYour keynot addedBilled to your key
- 6OpenRouterYour keynot addedBilled to your key
- 7OVHcloudYour keynot addedBilled to your key
- 8Ramp RouterYour keynot addedBilled to your key
- 9Vercel AI GatewayYour keynot addedBilled to your key
- 10SiliconFlowYour keynot addedBilled to your key
- 11NVIDIAYour keynot addedBilled to your key
- 12Nous ResearchYour keynot addedBilled to your key
- 13InfronYour keynot addedBilled to your key
- 14LLM GatewayYour keynot addedBilled to your key
- 15InferXYour keynot addedBilled to your key
- 16Ollama CloudYour keynot addedBilled to your key
- 17AI/ML APIYour keynot addedBilled to your key
- 18NanoGPTYour keynot addedBilled to your key
- 19FluenceYour keynot addedBilled to your key
- 20CortecsYour keynot addedBilled to your key
- 21HeliconeYour keynot addedBilled to your key
- 22PoeYour keynot addedBilled to your key
- 23Jiekou.AIYour keynot addedBilled to your key
- 24KenariYour keynot addedBilled to your key
Benchmarks
- GPQA Diamond
- 68.8%
- LiveCodeBench
- 77.7%
- AIME
- 89.3%
- HLE
- 9.8%
Quick start
curl
curl https://anyrouter.dev/api/v1/chat/completions \
-H "Authorization: Bearer $ANYROUTER_API_KEY" \
-H "Content-Type: application/json" \
-d '{
"model": "openai/gpt-oss-20b",
"messages": [
{
"role": "user",
"content": "Say hi in 3 words."
}
]
}'Details
- Owner
- OpenAI
- Kind
- Chat
- Released
- Apr 2025
- Context
- 131K
- Data policy
- Provider retention
- Endpoint
/api/v1/chat/completions
- Low-latency, local or specialized use
- Agentic tasks with function calls
- Fine-tuning on consumer hardware
Architecture and licenseKiến trúc và giấy phép架构与许可
- TypeLoại类型
- Mixture of experts[3]
- Training cutoffMốc dữ liệu训练截止
- 2024-06[3]
- LicenseGiấy phép许可
- Apache-2.0 · Open weightsTrọng số mở开放权重[1]
StrengthsĐiểm mạnh优势[1]
- Runs within 16 GB of memory with MXFP4 weights
- Reasoning effort set to low, medium or high
- Native function calling, web browsing and Python execution
Known limitsGiới hạn đã biết已知局限[3],[1]
- Hallucinates more than o4-mini on SimpleQA and PersonQA
- Raw chain of thought can hallucinate; filter it before showing users
- Works only with the harmony response format
API notesGhi chú APIAPI 说明[1]
- Reasoning level set in the system prompt, e.g. "Reasoning: high"
- Supports function calling and structured outputs
Links and versionsLiên kết và phiên bản链接与版本
- AnnouncementBài công bố发布公告
- openai.com
- Model cardThẻ mô hình模型卡
- huggingface.co
- PaperBài báo论文
- arxiv.org
SourcesNguồn来源
- [1]openai/gpt-oss-20b model cardaccessedtruy cập访问于 2026-10-14
- [2]gpt-oss-20b config.jsonaccessedtruy cập访问于 2026-10-14
- [3]gpt-oss-120b & gpt-oss-20b Model Cardaccessedtruy cập访问于 2026-10-14
Upstreams and pricingUpstream và giá上游与价格
Share cards25 images

























Uptime + latency
—recent checks · all routes
No health checks recorded for this model yet.
24h
No traffic in the last 24h.
Across all upstreams
Loading usage…