CAIS Text Capabilities Index: leaderboard
Composite CAIS AI Dashboard text index averaging Humanity's Last Exam, ARC-AGI-2, TextQuests, and SWE-bench Pro for models with all component scores.
Metric: Text Capabilities Index (imported). Source: dashboard.safe.ai. Source: benchmarklist.com. Saturation forecast: Around June 2027. 40 models tracked.
Top models
| # | Model | Score | Overall rank |
|---|---|---|---|
| 1 | GPT-5.5 | 55.9 | #31 |
| 2 | Claude Opus 4.8 | 53.8 | #37 |
| 3 | Gemini 3.1 Pro (Preview) | 52.9 | #58 |
| 4 | GPT-5.4 | 49.3 | #79 |
| 5 | Gemini 3.5 Flash | 48.9 | #60 |
| 6 | Claude Opus 4.7 | 46.9 | #49 |
| 7 | Claude Opus 4.6 | 44 | #63 |
| 8 | Gemini 3 Pro | 38.4 | #80 |
| 9 | Claude Opus 4.5 | 36.6 | #83 |
| 10 | Gemini 3 Flash | 35.6 | #99 |
| 11 | GPT-5.2 | 33.8 | #110 |
| 12 | Kimi K2.6 | 32.8 | #104 |
| 13 | Claude Sonnet 4.6 | 32.6 | #90 |
| 14 | Grok 4.20 | 32.5 | #175 |
| 15 | DeepSeek V4 Pro | 32.1 | #102 |
No result here: #3 Claude Opus 5.5, #6 GPT-6 Astra, #7 GPT-6.1 Sol.
Interactive version: theaggregate.ai/benchmark?slug=cais-text-capabilities-index · How It Works · Data refreshed daily, snapshot 2026-10-20.