UniQL - Trino: leaderboard

Metric: Execution accuracy (%) on the Trino dialect of UniQL (1,534 BIRD development-set questions with human-verified SQL in 16 aligned dialects): given the target dialect, schema and question, the model writes one query whose execution result must match the gold result; inference-only, one run; higher is better. Source: arxiv.org. Saturation forecast: Around 2028. 13 models tracked.

Top models

#ModelScore
1Claude Sonnet 4.556.06
2Gemini 2.5 Pro55.61
3GPT-5.1 Codex50.72
4GPT-5 Mini49.02
5DeepSeek V4 Flash47.59
6Qwen 3 32B46.61
7Qwen 3 8B44.2
8Qwen 3 4B41.46
9Llama 3 70B Instruct39.05
10GPT-3.5 Turbo35.85
11Qwen 3 1.7B34.16
12Llama 3 8B Instruct22.75

Interactive version: theaggregate.ai/benchmark?slug=uniql-trino · How It Works · Data refreshed daily, snapshot 2026-09-29.