Expired75% confidenceFactTime unknown
GPT-OSS 120B surpasses the performance of the Qwen3 235B AR2B reasoning model with less than half the parameter count and hardware cost.
1
Sources
75%
Confidence
Medium-term (~90 days)
Relevance
6/1/2026
First Seen
This claim has expired. It is kept for the record and no longer describes the current state.
Valid until: 8/30/2026(expired)
Sources
Related Entities
Related Claims
UnverifiedGPT-OSS 120B需要大于80GB显存（如H100 GPU）才能运行73% similarUnverifiedGPT-OSS 120B performance is on par with O4 mini.71% similarUnverifiedGPT-OSS-120B officially requires no less than 60GB of VRAM.70% similarUnverifiedThe most advanced HBM3e memory bandwidth is approximately 4.8 TB/s, yet GPU compute utilization during inference often reaches only 30%–60%.67% similarUnverifiedDeploying an open-source model with 70 billion parameters running at FP16 precision requires at least four A100/H100 GPUs with 80 GB of VRAM each, with hardware procurement costs alone exceeding one million yuan.66% similar
Cite This Claim
Stable URI
https://kongchang.com/claim/23214API
curl https://kongchang.com/api/v1/knowledge/claims/23214MCP
get_claim(id=23214)