Reasoning benchmark

ARC-AGI-3 Provider Adapter leaderboard

ARC-AGI-3 Semi-Private: Provider Adapter harness. Every model the catalog carries a published ARC-AGI-3 Provider Adapter value for, ranked by that value.

CategoryReasoning

Published byARC Prize verified ARC-AGI-3 results

ARC Prize verified interactive reasoning results using a provider-specific stateful adapter, kept separate from Standard harness results.

ARC-AGI-3 Provider Adapter ranking

8 models with a published ARC-AGI-3 Provider Adapter value, ordered by that value, highest first. Models the source has not scored on this benchmark are not listed — they are unmeasured, not last.

Models ranked by their published ARC-AGI-3 Semi-Private: Provider Adapter harness value
Rank	Model	Provider	ARC-AGI-3 Provider Adapter
1	GPT-6 Astra	OpenAI	98.55
2	Gemini 4 Argon	Google	97.30
3	GPT-6.1 Sol	OpenAI	96.18

About ARC-AGI-3 Provider Adapter

What the benchmark measures and how its value is published, from the source’s own record.

What is ARC-AGI-3 Provider Adapter?

ARC-AGI-3 Provider Adapter (ARC-AGI-3 Semi-Private: Provider Adapter harness) is a reasoning benchmark published by ARC Prize verified ARC-AGI-3 results. ARC Prize verified interactive reasoning results using a provider-specific stateful adapter, kept separate from Standard harness results.

How is ARC-AGI-3 Provider Adapter scored?

One value per model is published by ARC Prize verified ARC-AGI-3 results. ModelScale prints that value to 2 decimal places, exactly as published, and never converts or rescales it. This page ranks the 8 models that carry a published value, highest value first; a model the source never scored is left out rather than ranked at zero. The source does not state whether a higher value is the better result, so neither does this page.

All leaderboards · Reasoning capability leaderboard

Explore models

Top 1-5

Top 6-10

Featured

Capabilities

Benchmarks

Compare

Featured model comparisons

Featured provider comparisons

Articles