News Summary

• ARC Prize published both numbers on the day GPT-6 Astra launched.
• It printed a full table of every reasoning level it ran.
• The difference is the harness.
• A harness is the software around a model.
• It sets the tools the model can reach, what it remembers between requests, and how its context gets managed.
• Run the same model through the benchmark’s own software, and it scores 62.7%.
• The gap is not a rounding error or a rival's complaint.
• It comes from the organisation that built the test.