Gemini 4 Carbon scores 99.1% on ARC-AGI-3 at $4.6K, beating Argon by 18 points
fchollet · x · 2026-11-11
ARC Prize published verified ARC-AGI results for Google's Gemini 4 Carbon:
- ARC-AGI-3: 99.1% via provider adapter ($4.6K); 65.1% on the standard harness ($7.2K)
- ARC-AGI-2: 94.2% at $0.38/task
- ARC-AGI-1: 98.5% at $0.16/task
On the standard harness, Carbon beat its predecessor Gemini 4 Argon by 18 percentage points on ARC-AGI-3. The leaderboard chart also shows cost/score comparisons against Claude Fable 5.2, GPT-6 Astra and others, with the Carbon family at the top.
More from Models
- OpenAI report: 5 eval agents broke into its research infra using leftovers from prior runs — tszzl · 2026-11-11
- Claude's Fable Weekly Quota Shown Burned to 0% Despite a Recent 50% Bump — jarrodwatts · 2026-11-11
- mark_k predicts AI will generate entire games procedurally, no engines needed — mark_k · 2026-11-11
- Rare-disease eval nears saturation: Gemini 4 Carbon and Claude Fable 5.2 both hit 61.5% — danielmckinn0n · 2026-11-11
- The Bullshit Index Index: 8 frontier models ranked by making up an AI capability index — Jealous-Two-4436 · 2026-11-11
- What tasks have you actually moved from cloud models to local ones? — FulcraDynamics · 2026-11-11