Expired80% confidenceFactTime unknown
Opus 4 outperformed all models in software engineering benchmarks, including Gemini 2.5 Pro.
1
Sources
80%
Confidence
Medium-term (~90 days)
Relevance
6/1/2026
First Seen
Valid until: 8/30/2026(expired)
Sources
Related Entities
Related Claims
UnverifiedThe four AI models tested were Claude Opus 4.5, Gemini 3 Pro, GLM 4.7, and Minimax M2.1.75% similarVerifiedClaude Opus 4.5 leads Gemini 3 Pro on the ARC-AGI-V2 benchmark by more than 6 percentage points.74% similarUnverifiedIn graduate-level reasoning and Agent tool-use benchmarks, Opus 4 is largely in the lead, though Gemini 2.5 Pro comes very close on some items.72% similarUnverifiedThe latest version of Gemini outperforms the older version of Claude on almost all benchmarks, with the sole exception of coding.70% similarUnverifiedThis article's tests conduct a horizontal comparison of Gemini 3.1 Pro with Claude Opus 4.6 and GPT 5.3 Codex68% similar
Cite This Claim
Stable URI
https://kongchang.com/claim/18600API
curl https://kongchang.com/api/v1/knowledge/claims/18600MCP
get_claim(id=18600)