MODEL / OPENAI
gpt-oss-120b
OpenAI targets one 80 GB accelerator. Correct serving requires Harmony format support.
一手来源总参数117B5.1B active / token
估算内存~80 GBMXFP4
上下文128KText
LicenseApache-2.0Primarily English
01 / 适合
- reasoning
- agents
- datacenter serving
02 / 估算适配
03 / 部署方式
vLLMAdvanced
vllm serve openai/gpt-oss-120b --dtype autoTensorRT-LLMAdvanced
trtllm-serve openai/gpt-oss-120bllama.cppMedium
llama-server -m /path/to/verified-model.gguf --port 8080注意
OpenAI targets one 80 GB accelerator. Correct serving requires Harmony format support.