Qwen 2.5 Turbo — benchmark results
Alibaba's fast, low-cost proprietary Qwen2.5 API tier, whose Turbo update (November 2024) extended context to 1M tokens via sparse attention. Provider: Alibaba. Released 2024-09-19. Access: API.
Unified ELO 1407 ± 29, rank #1193 of 1776 rated models, from 11 benchmark results.
Strongest benchmark results
| Benchmark | Score | Metric | Percentile |
|---|---|---|---|
| AA MATH-500 | 80.53 | Accuracy (%) | 47.3 |
| AI Chess Leaderboard (Reasoning) | 574 | Elo | 30.9 |
| AA MMLU-Pro | 63.31 | Accuracy (%) | 24.7 |
| AI Chess Leaderboard (Continuation) | 438 | Elo | 23.5 |
| BenchTable | 25.1 | Total Score (%) | 20.3 |
| Artificial Analysis Intelligence Index | 6.3 | Intelligence Index | 18.3 |
| AA GPQA Diamond | 41.01 | Accuracy (%) | 18 |
| AA LiveCodeBench | 16.3 | Pass@1 (%) | 17 |
| AA Humanity's Last Exam | 4.21 | Accuracy (%) | 15.6 |
| AA SciCode | 15.28 | Accuracy (%) | 14.9 |
| ResearchCodeBench | 8 | Task Success Rate (%) | 3.2 |
Interactive version: theaggregate.ai/model?slug=qwen-2-5-turbo · How the rankings work · Data refreshed daily, snapshot 2026-07-22.