Unverified50% confidenceBenchmarkExact time
In the five-part image-recognition test, Gemini 4 Argon got all five correct, tied with Opus, while GPT got four correct; in the clock-reading question, GPT had already read 28 to 29 minutes but ultimately answered "can't see clearly," while Argon reported 10:29.
1
Sources
50%
Confidence
Short-term (~14 days)
Relevance
10/14/2026
First Seen
Valid until: 10/28/2026
Sources
Related Entities
Related Claims
UnverifiedGemini 4 Argon and GPT-6 Sol tie on FrontierCode 1.1 and beat GPT-6.1 Sol, indicating that the gap in coding ability among current top models has narrowed to a very small range.70% similarUnverifiedA Bilibili uploader used the same prompt to have Gemini 4 Argon, GPT-6 Astra, and Claude Opus 5.5 each generate a video with code for a comparative hands-on test.68% similarUnverifiedOn purely agentic coding and computer operation tasks, GPT-6.1 SoC still slightly outperforms Gemini 4 Argon in certain specific scenarios.66% similarUnverifiedIn the 3D Monument to the People's Heroes agentic programming test, GPT-6 Luna took 7 minutes and 15 seconds and was the only model to produce the inscription on the monument's body, but rendered the vertical layout as horizontal.66% similar
Cite This Claim
Stable URI
https://kongchang.com/claim/1030285API
curl https://kongchang.com/api/v1/knowledge/claims/1030285MCP
get_claim(id=1030285)