Qwen 3 32B (Thinking): benchmark results
Qwen 3 32B evaluated with thinking enabled. Provider: Alibaba. Released 2025-04-29. Access: Open.
Aggregate Rating 1574 ± 6, rank #780 of 2168 rated models, from 519 benchmark results.
Strongest benchmark results
| Benchmark | Score | Metric | Percentile |
|---|---|---|---|
| ESF-Bench - Multi-Source | 87.3 | Tagged slot accuracy (%; slots tagged with a MS (Multi-Sourc | 100 |
| ESF-Bench - Personalization | 66.8 | Tagged slot accuracy (%; slots tagged with a PER (Personaliz | 100 |
| ESF-Bench - Relative | 91.2 | Tagged slot accuracy (%; slots tagged with a REL (Relative) | 100 |
| FinEval 6.0 - Agent Application | 84.62 | Score (0-100) | 100 |
| LLM Stats (Codeforces Rating - Qwen3 CodeElo) | 1977 | Codeforces Rating | 100 |
| RFEval | 73.29 | Contrast-conditional reasoning faithfulness (%): after a cou | 100 |
| RFEval - Logical Reasoning | 88.62 | Contrast-conditional reasoning faithfulness (%): after a cou | 100 |
| RFEval - Mathematical Reasoning | 47.87 | Contrast-conditional reasoning faithfulness (%): after a cou | 100 |
| RFEval - Table Reasoning | 89.84 | Contrast-conditional reasoning faithfulness (%): after a cou | 100 |
| TeleCom-Bench - Wireless Network | 73.04 | LLM-judge score (%; open-ended wireless-network product-know | 100 |
| Swallow - English MT-Bench - Math | 99.7 | Judge Score (normalized, %) | 98.5 |
| SEA-HELM (English) - LiveCodeBench v6 | 69.34 | Normalized Score | 94.7 |
Interactive version: theaggregate.ai/model?slug=qwen-3-32b-thinking · How It Works · Data refreshed daily, snapshot 2026-10-20.