Tinybird AI SQL Benchmark - Exactness: leaderboard
Metric: Result exactness vs human reference queries (0-100). Source: llm-benchmark.tinybird.live. 327 models tracked.
Top models
| # | Model | Score |
|---|---|---|
| 1 | Claude Opus 4.7 | 65.48 |
| 2 | Claude Fable 5 | 65.05 |
| 3 | Claude Fable 5.1 | 61.53 |
| 4 | Gemini 3 Pro (Preview) | 60.3 |
| 5 | Qwen 3.5 Plus | 60.17 |
| 6 | Gemini 3.5 Flash | 60.05 |
| 7 | Grok 4.6 | 59.71 |
| 8 | Gemini 3.6 Flash | 59.48 |
| 9 | Gemini 2.5 Pro (Preview 05-06) | 58.94 |
| 10 | Gemini 3 Flash (Preview) | 58.35 |
| 11 | Gemini 2.5 Pro | 57.98 |
| 12 | Claude Opus 4.8 | 57.77 |
| 13 | GLM-5.3 Flash | 57.62 |
| 14 | Seed 2.0 Lite | 57.35 |
| 15 | Muse Spark 1.2 | 57.22 |
Interactive version: theaggregate.ai/benchmark?slug=tinybird-ai-sql-benchmark-exactness · How It Works · Data refreshed daily, snapshot 2026-09-19.