Model Releases
About a year ago, before it was on anyone's radar, we began exploring the idea of a benchmark for open-ended invention. Since then, we've de…
About a year ago, before it was on anyone's radar, we began exploring the idea of a benchmark for open-ended invention. Since then, we've developed several promising directions that will serve as the
About a year ago, before it was on anyone's radar, we began exploring the idea of a benchmark for open-ended invention. Since then, we've developed several promising directions that will serve as the foundation for ARC 4 and ARC 5. We're incredibly excited to share what we've been building. We're still on track to release ARC 4 in Q1 next year, as promised. ARC-AGI-4 will be a benchmark for autonomous open-ended innovation. It will continue our commitment to open-source, giving the research community a shared target for progress that benefits all of humanity. Despite rapid model progress, humans still significantly outperform AI at …
Related
- If your benchmark relies on a static dataset or sampling from a static distribution densely known at training time, then it is fundamentally…
- Standard RL benchmarks are episodic and stationary, so they don't capture the the characteristics of real-world deployment. Morpheus is a ne…
- GPT-5.5 on ARC-AGI (Verified) ARC-AGI-2: - Max: 85.0%, 1.87 - High: 83.3%, 1.45 - Med: 70.4%, 0.86 - Low: 33%, 0.35 GPT-5.5 is now state…
- Reporting benchmark results as a scalar number, e.g. '75% on XYZ' is completely meaningless at this point. You should always report efficien…
Source: Francois Chollet (X) | 2026-09-12