Expired50% confidenceOpinionExact time
Steinfurt considers Gemini 3 Pro to be the best-performing model on chess tasks he has seen, and the team speculates that Google carried out in-depth chess-specific post-training.
1
Sources
50%
Confidence
Medium-term (~90 days)
Relevance
7/12/2026
First Seen
Valid until: 10/10/2026(expired)
Sources
Related Claims
UnverifiedBefore Gemini 3 Pro, the best-performing model on chess tasks was Grok 4.77% similarUnverifiedGoogle has launched the Gemini 3.1 Pro model, bringing significant improvements in abstract reasoning and multimodal tasks.68% similarUnverifiedClaude Code supports integration with third-party models, including domestic large models and Google Gemini.64% similar
Cite This Claim
Stable URI
https://kongchang.com/claim/490598API
curl https://kongchang.com/api/v1/knowledge/claims/490598MCP
get_claim(id=490598)