Opus 5 reaches about 30% on ARC-AGI-3 as GPT-5.6 trails behind
haider1 · x · 2026-07-28
A post on the ARC Prize leaderboard suggests Opus 5 has jumped to about 30% on ARC-AGI-3, far above GPT-5.6’s earlier single-digit result.
The author argues that if this pace holds, ARC-AGI-3 could saturate in just 3–4 months, especially with GPT-6 coming soon and likely pushing the frontier further. The screenshot also shows Opus 5 leading ARC-AGI-1 and ARC-AGI-2, while GPT-5.6 variants remain behind on ARC-AGI-3.
More from Models
- Users discuss what they actually use Opus 5 for beyond coding — remilouf · 2026-07-29
- Anthropic Hints at Achieving Recursive Self-Improvement, Calls for Pacing Frontier — daniel_mac8 · 2026-07-29
- Claude Opus 5 tops DeepSWE with a 74% score and a claimed 28% cost edge — daniel_mac8 · 2026-07-29
- LiquidAI’s 230M LFM2.5 encoder trends on Hugging Face — LiquidAI · 2026-07-29
- Anthropic may be 1.5 generations ahead internally, with Fable 5.1 weeks away — haider1 · 2026-07-29
- Moonshot’s Kimi K3 is a 2.8T open-weight MoE model with 1M-token context — alex_verem · 2026-07-29