Feed Entities Podcasts Deep Dives Developers About

[KongchangAI]

Real-time AI knowledge base. Every claim is traceable to its source, with data freshness under 24 hours.

Powered by 240+ first-hand AI sources

Knowledge

Feed
Hot Entities
Podcasts

Developers

CLI
MCP Server
REST API
About

© 2026 KongchangAI. The freshest, most accurate AI knowledge

Unverified75% confidenceFactTime unknown

GPT-OSS 120B以不到一半的参数量和硬件成本超越了千问3 235B AR2B推理模型的性能

1

Sources

75%

Confidence

Medium-term (~90 days)

Relevance

6/1/2026

First Seen

Valid until: 8/30/2026

Sources

OpenAI开源GPT-OSS：16G显存跑O4级模型，部署教程全解析

bilibili九天Hector

Related Entities

GPT-OSS 120Bmodel 千问3 235Bmodel

Related Claims

UnverifiedGPT-OSS 120B需要大于80GB显存（如H100 GPU）才能运行73% similar UnverifiedGPT-OSS 120B性能与O4 mini不相上下71% similar UnverifiedGPT-OSS-120B官方要求显存不低于60GB70% similar Unverified最先进的HBM3e内存带宽约4.8TB/s，推理过程中GPU算力利用率往往只有30%-60%67% similar Unverified部署700亿参数规模的开源模型以FP16精度运行至少需要4张80GB显存的A100/H100 GPU，仅硬件采购成本超过百万人民币66% similar

Cite This Claim

Stable URI

https://kongchang.com/claim/23214

API

curl https://kongchang.com/api/v1/knowledge/claims/23214

MCP

get_claim(id=23214)