Reasoning Benchmark

ARC-AGI (v2)

ARC-AGI-2 abstraction-and-reasoning benchmark (arcprize.org/arc-agi/2, dataset repo github arcprize/ARC-AGI-2). v2 half of the card's 'ARC-AGI v1 / v2' row.

Category

Reasoning

Score type

percent

Scale

100%

higher is better

Models tested

2

What is ARC-AGI (v2)?

ARC-AGI (v2) is a benchmark for general reasoning and comprehension. The official description says: "ARC-AGI-2 abstraction-and-reasoning benchmark (arcprize.org/arc-agi/2, dataset repo github arcprize/ARC-AGI-2)". Below we describe in simple words what benchmarks of this type generally examine.

What does it test?

ARC-AGI (v2) measures general reasoning and comprehension. In detail, it checks:

  • drawing the right conclusion from a passage or scenario
  • commonsense judgement about everyday situations and words
  • counting, comparing and reasoning over several facts at once
  • avoiding shortcuts that memorise answers instead of reasoning

How to read the score

The score is accuracy: the percentage of tasks answered or solved correctly. 100% means every task was solved perfectly; higher is better.

Official methodology note: ARC-AGI-2 abstraction-and-reasoning benchmark (arcprize.org/arc-agi/2, dataset repo github arcprize/ARC-AGI-2). v2 half of the card's 'ARC-AGI v1 / v2' row.

Frequently asked questions

What is the ARC-AGI (v2) benchmark?

ARC-AGI (v2) is a benchmark for general reasoning and comprehension. The official description says: 'ARC-AGI-2 abstraction-and-reasoning benchmark (arcprize.org/arc-agi/2, dataset repo github arcprize/ARC-AGI-2)'. Below we describe in simple words what benchmarks of this type generally examine.

What does ARC-AGI (v2) test?

ARC-AGI (v2) mainly measures general reasoning and comprehension. Concretely, it checks: drawing the right conclusion from a passage or scenario; commonsense judgement about everyday situations and words; counting, comparing and reasoning over several facts at once.

How is the ARC-AGI (v2) score calculated?

The score is accuracy: the percentage of tasks answered or solved correctly. 100% means every task was solved perfectly; higher is better. The declared maximum score of this benchmark is 100.0.

Where do I get ARC-AGI (v2) data and scores?

The official source is https://arcprize.org/arc-agi/2. The official benchmark page/reference is linked on this page. Every score on Open LLM Radar is also available as open data: download the scores JSON (http://open-llm-radar-api:8000/api/v1/benchmarks/arc-agi-v2/scores) or pull it directly from the Open LLM Radar API with any HTTP client.

Model scores 2 models tested

Reset
# Model Score Date Source Verified
1 Inkling 2 Mini
Thinking Machines
49.3%
20.11.2026 Report Verified
2 Qwen3.8-27B
Qwen Team (Alibaba Cloud)
42.4%
20.11.2026 Report Unverified

What now?

You've seen how models score on ARC-AGI (v2) — here's what to do next.