Vals Index: leaderboard
This board is benchmarklist.com's copy of the Vals Index, a composite that weights sector averages of Vals AI benchmark scores by each sector's share of US GDP, not a single task. Its rows were imported from Vals AI on 28 July 2026, before the August 2026 v2 revision restored a legal sector, so they cover finance and coding only. The number is a percentage, higher is better. Source: vals.ai/benchmarks/vals_index.
Metric: Score (imported). Source: vals.ai. Source: benchmarklist.com. Saturation forecast: Around August 2027. 40 models tracked.
Top models
| # | Model | Score | Overall rank |
|---|---|---|---|
| 1 | Claude Fable 5 (Max) | 75.14 | #16 (Claude Fable 5) |
| 2 | Claude Opus 5 | 74.82 | #12 |
| 3 | Kimi K3 | 74.7 | #30 |
| 4 | GPT-5.6 Sol (Max) | 73.12 | #19 (GPT-5.6 Sol) |
| 5 | Claude Opus 4.8 (Max) | 70.36 | #37 (Claude Opus 4.8) |
| 6 | GPT-5.6 Luna (Max) | 69.88 | #68 (GPT-5.6 Luna) |
| 7 | Claude Sonnet 5 (Max) | 68.61 | #57 (Claude Sonnet 5) |
| 8 | Muse Spark 1.1 (xHigh) | 68.41 | #52 (Muse Spark 1.1) |
| 9 | GPT-5.5 (xHigh) | 67.95 | #31 (GPT-5.5) |
| 10 | Claude Opus 4.7 (High) | 66.1 | #49 (Claude Opus 4.7) |
| 11 | Grok 4.5 (High) | 65.3 | #48 (Grok 4.5) |
| 12 | GPT-5.6 Terra (xHigh) | 65.14 | #41 (GPT-5.6 Terra) |
| 13 | GLM-5.2 (Max) | 65.02 | #73 (GLM-5.2) |
| 14 | Gemini 3.5 Flash (High) | 62.75 | #60 (Gemini 3.5 Flash) |
| 15 | Gemini 3.6 Flash (High) | 62.43 | #51 (Gemini 3.6 Flash) |
No result here: #3 Claude Opus 5.5, #6 GPT-6 Astra, #7 GPT-6.1 Sol.
Interactive version: theaggregate.ai/benchmark?slug=vals-index · How It Works · Data refreshed daily, snapshot 2026-10-20.