benchmark-result:3515
o3 on ARC-Challenge
CodeSOTA verified- Model
- o3
- Benchmark
- ARC-Challenge
- Metric
- accuracy
- Value
- 98.1
Verified o3 ARC-Challenge row.
Registry note: 0-shot. Source: OpenAI simple-evals (2025).
benchmark-result:3516
o4-mini on ARC-Challenge
CodeSOTA verified- Model
- o4-mini
- Benchmark
- ARC-Challenge
- Metric
- accuracy
- Value
- 97.3
Verified o4-mini ARC-Challenge row.
Registry note: 0-shot. Source: OpenAI simple-evals (2025).
benchmark-result:3517
Gemini 2.5 Pro on ARC-Challenge
CodeSOTA verified- Model
- Gemini 2.5 Pro
- Benchmark
- ARC-Challenge
- Metric
- accuracy
- Value
- 97.8
Verified Gemini 2.5 Pro ARC-Challenge row.
Registry note: 0-shot CoT. Source: Gemini 2.5 Pro technical report (April 2025).