Unverified50% confidenceReleaseExact time
Google has opened Gemini 4 Argon to all subscribers, and the model scores 97.5% on the ARC AGI three evaluations.
1
Sources
50%
Confidence
Short-term (~14 days)
Relevance
10/18/2026
First Seen
Valid until: 11/1/2026
Sources
AI Weekly: DeepSeek's New Performance Baseline, Gemini 4 Cuts Costs 40%, Kling 4.0 Single-Shot Video
bilibili小行星AI观测站10/18/2026
Related Entities
Related Claims
UnverifiedGoogle reports that Gemini 4 Argon achieved a score of 77.9% on the DeepSWE 1.1 software engineering benchmark (as self-reported by the vendor)77% similarUnverifiedGoogle's official post only disclosed the output limit for Gemini 4 Argon, not the input context window; the figure of 1 million input tokens comes from Artificial Analysis rather than official confirmation.77% similarUnverifiedGoogle reports Gemini 4 Argon at 55.0% on FrontierSWE v2, trailing GPT-6 Astra by approximately 10.5 percentage points; on Terminal-Bench 4.0 it scores 57.4%, trailing Opus 5.5 by nine percentage points.76% similarUnverifiedGoogle provides a free Search grounding quota for Gemini 4 Argon: 5,000 free requests per month (shared by Gemini 3 and subsequent models), and $14 per 1,000 requests afterward.76% similarUnverifiedAmong the 19 benchmarks Google published, Gemini 4 Argon ranks first on 14, including a Wells index of 68.9 for knowledge work, 84.2 for reasoning over million-token graph data, and 91.7 for long-video understanding.75% similar
Cite This Claim
Stable URI
https://kongchang.com/claim/1059823API
curl https://kongchang.com/api/v1/knowledge/claims/1059823MCP
get_claim(id=1059823)