Unverified50% confidenceBenchmarkExact time
A full V2 training run takes approximately 5 hours on a consumer GPU, ultimately achieving approximately 74% top-1 and 86% top-5 accuracy on unseen text.
1
Sources
50%
Confidence
Long-term
Relevance
10/11/2026
First Seen
Sources
Related Entities
Related Claims
VerifiedTraining a model at the level of GPT-4 requires tens of thousands of A100/H100 GPUs working in concert for several months.69% similarUnverifiedIndustry estimates suggest GPT-4's training used approximately 25,000 NVIDIA A100 GPUs over several months, with compute costs alone exceeding $100 million67% similarUnverifiedGPT-Live Mini and Live Medium achieved scores in the 70%/80% range on GPQA, essentially reaching GPT-5-level intelligence.64% similarUnverifiedGPT-4o mentioned in the paper achieved over 85% accuracy in almost all tested scenarios.64% similar
Cite This Claim
Stable URI
https://kongchang.com/claim/997214API
curl https://kongchang.com/api/v1/knowledge/claims/997214MCP
get_claim(id=997214)