Unverified50% confidenceFactExact time
Pavel Hurin's follow-up test found that running the same benchmark on the anti-gravity CLI rather than the Gemini CLI, Argon jumped from "decent" to frontier level, "very close to Opus 5.5 Max", and he ran it three times to confirm the result.
1
Sources
50%
Confidence
Short-term (~14 days)
Relevance
10/14/2026
First Seen
Valid until: 10/28/2026
Sources
Related Entities
Related Claims
VerifiedGemini 4 Argon achieved a new record of 77.9% on the Deep7e long-workflow software engineering benchmark, surpassing Claude Opus 5.5.67% similarUnverifiedDeepMind's official blog post stated that Gemini 4 Argon surpasses all other frontier models on the ERQA benchmark.67% similarPartially VerifiedGemini 4 Argon topped 13 tests, and tied for first in one.67% similarUnverifiedDeveloper Pavel Hurin tested Argon with a bug hunter task, and the result was comparable to Muse Spark 1.3, clearly lagging behind Astra, Fable and Opus.66% similarUnverifiedGemini 4 Argon did not take first place in all tests; in complex engineering, terminal operations, and scientific research tasks, other models each led in some areas.65% similar
Cite This Claim
Stable URI
https://kongchang.com/claim/1032319API
curl https://kongchang.com/api/v1/knowledge/claims/1032319MCP
get_claim(id=1032319)