AIMultiple - Text-to-SQL (Adjudicated): leaderboard
AIMultiple Text-to-SQL adjudicated score: misses under strict execution match are re-read by a blind three-model jury from other model families, crediting equivalent answers and questions whose BIRD gold query is broken. Measured over each model's correctly routed questions; an optimistic LLM-jury estimate, %, higher is better.
Metric: Execution match with jury-credited equivalents and broken golds (%, over correctly routed questions). Source: aimultiple.com. Saturation forecast: Around February 2027. 81 models tracked.
Top models
| # | Model | Score | Overall rank |
|---|---|---|---|
| 1 | Claude Opus 5.5 | 91.1 | #3 |
| 2 | Claude Sonnet 5.5 | 83.2 | #14 |
| 3 | Gemini 3.6 Flash | 83.1 | #57 |
| 4 | Claude Fable 5.1 | 82.8 | #10 |
| 5 | Claude Opus 5 | 82.4 | #16 |
| 6 | Gemini 3.8 Flash | 82.2 | #30 |
| 7 | GPT-6.1 Sol | 81.8 | #7 |
| 8 | Gemini 4 Argon | 81.4 | #9 |
| 9 | GPT-6 Astra | 80.4 | #6 |
| 10 | Claude Fable 5 | 80.3 | #20 |
| 11 | Muse Wonder | 79.8 | #18 |
| 12 | GPT-6.1 Pro Sol | 79.7 | #11 |
| 13 | Kimi K3.1 Pro | 79.2 | #17 |
| 14 | Gemini 3.7 Flash | 79.1 | #24 |
| 15 | GPT-5.6 Sol | 78.5 | #22 |
Interactive version: theaggregate.ai/benchmark?slug=aimultiple-text-to-sql-adjudicated · How It Works · Data refreshed daily, snapshot 2026-11-05.