Capability
Interactive Agent Reasoning
Score
30.16/100
- Confidence
- 65%
- Evidence records
- 1
Versioned model-family evidence
Evidence preview · noindex
This profile exposes only reviewed identity, configuration and score facts from the current release. It does not yet contain approved model-specific prose.
Current immutable release
A precise 0–100 score is shown only where reviewed evidence supports one. Confidence describes evidence strength separately; an insufficient-evidence row is not a zero.
Capability
Score
30.16/100
Capability
Score
Insufficient evidence
Score
Insufficient evidence
Task
Score
Insufficient evidence
Identity and provenance
claude-opus-5
Configuration configuration-claude-opus-5-high · Representative
Reviewed 8 Aug 2026
Selected the highest reviewed, published ARC-AGI-3 RHAE observation for this source family; equal scores resolve by configuration slug ascending.
5b6e0703ea86e50b3cb7bae331add0f8ffbb28b003698d3af5b1c0decccdccb5
Traceable inputs
arc-agi-3-official-verified
Metric: arc-agi-3-rhae