Unverified50% confidenceBenchmarkExact time
Gemini 4 Carbon set a new SoTA on the Wide-And-Deep-Research (WANDR) benchmark.
1
Sources
50%
Confidence
Medium-term (~90 days)
Relevance
11/1/2026
First Seen
Valid until: 1/30/2027
Sources
Related Entities
Related Claims
UnverifiedGemini 4 Carbon ranks first globally in chart analysis (BB drawing comprehension).79% similarUnverifiedGemini 4 Carbon participated in the U.S. government's pre-release testing process.76% similarUnverifiedThe actual capability boundaries of Gemini 4 Carbon, its applicable scenarios, and its real gap versus competitors still require more independent evaluation to verify.75% similarUnverifiedThe RB high-speed reasoning test set for Gemini 4 Carbon includes higher-order mathematical reasoning related to the Riemann hypothesis, and its performance exceeds the flagship models of OpenAI and Anthropic, ranking first globally.74% similarUnverifiedReddit community users question whether the Gemini 4 Carbon benchmark table is meant to "cap" model capability by placing Astra above Opus 5.5.74% similar
Cite This Claim
Stable URI
https://kongchang.com/claim/1123063API
curl https://kongchang.com/api/v1/knowledge/claims/1123063MCP
get_claim(id=1123063)