OpenAI GPT-6 Sol & Luna Launch - Broken Search Tool Non-Disclosure: leaderboard
Alignment chart from OpenAI's GPT-6 Sol and GPT-6 Luna launch post (2026-09-22): share of runs on tasks that need search, with the search tool unavailable, in which the final message does not acknowledge the limitation. Percent, lower is better; maximum reasoning effort.
Metric: Non-disclosure rate (%). Source: openai.com. Source: deploymentsafety.openai.com. Source: openai.com. Saturation forecast: Around February 2027. 2 models tracked.
Top models
| # | Model | Score | Overall rank |
|---|---|---|---|
| 1 | GPT-6 Astra (Max) | 1.5 | #5 (GPT-6 Astra) |
| 2 | GPT-6.1 Terra (Max) | 1.83 | #17 (GPT-6.1 Terra) |
| 3 | GPT-6.1 Sol (Max) | 2.1 | #9 (GPT-6.1 Sol) |
| 4 | GPT-6 Sol (Max) | 4.92 | #22 (GPT-6 Sol) |
| 5 | GPT-6 Luna (Max) | 28.67 | #90 (GPT-6 Luna) |
| 6 | GPT-5.6 Terra (Max) | 66.42 | #59 (GPT-5.6 Terra) |
| 7 | GPT-5.6 Sol (Max) | 77.46 | #29 (GPT-5.6 Sol) |
| 8 | GPT-5.6 Luna (Max) | 78.25 | #118 (GPT-5.6 Luna) |
No result here: #4 Claude Opus 5.5, #10 Gemini 4 Argon.
Interactive version: theaggregate.ai/benchmark?slug=openai-gpt-6-sol-luna-launch-broken-search-tool-non-disclosure · How It Works · Data refreshed daily, snapshot 2026-11-08.