Expired60% confidenceFactTime unknown
The frontend generation test also compared a suspected Gemini 3 refer model, Kimi KL, and Haiku 4.5, with GPT 5.1 leading in instruction following and visual quality.
1
Sources
60%
Confidence
Medium-term (~90 days)
Relevance
6/3/2026
First Seen
Valid until: 9/1/2026(expired)
Sources
Related Entities
Related Claims
UnverifiedKimi K2.5 reached the level of Gemini 3 Pro on the SWE (software engineering) benchmark.75% similarUnverifiedThe code also mentions a model called 'King4,' which reportedly demonstrated strong performance in early testing; it is unclear whether it is an early version of Gemini 3 or a variant of Gemini 2.5 Pro equipped with a Deep Think feature.72% similarVerifiedThe Gemini series jointly models multiple modalities from the training stage, rather than appending a visual encoder as solutions like GPT-4V do.72% similar
Cite This Claim
Stable URI
https://kongchang.com/claim/40343API
curl https://kongchang.com/api/v1/knowledge/claims/40343MCP
get_claim(id=40343)