Anthropic is using ARC-AGI-3 to gauge Opus 5’s novel problem-solving ability
typewriters · x · 2026-07-25
Anthropic is reportedly using ARC-AGI-3 to measure the novel problem-solving ability of Opus 5.
The attached chart compares models by total evaluation cost on the x-axis and score on the y-axis. In the visual, Opus 5 (high) is shown around 30% on ARC-AGI-3, while Opus 4.8 (high) sits near the low single digits and GPT-5.6 Sol is plotted below 10% at higher cost.
The post frames ARC-AGI-3 as a benchmark for frontier-style novel problem solving rather than routine benchmark polishing.
Related event: Anthropic Releases Claude Opus 5: SOTA Performance at Half the Price(128 posts)→
More from Models
- Looking for a classifier of software engineering task shapes to pick models per task — StewartalsopIII · 2026-09-11
- DeepSeek V4 Pro API to continue after Sept 2026, billing unchanged — teortaxesTex · 2026-09-11
- DeepSeek V4.1 Flash Hits 98% of GPT-6 Astra's Score at 1.4% of the Cost in Third-Party Benchmark — ayushtweetshere · 2026-09-11
- TheZvi Polls: Has Your Coding Model Choice Changed Since Fable 5.1 and Astra? — TheZvi · 2026-09-11
- antirez Weighs In on Anthropic Banning Minors From Using Claude — antirez · 2026-09-11
- Engram's random reads don't suit SSDs; CPU-memory over NVLink could serve all 72 GPUs — bookwormengr · 2026-09-11