Qwen QwQ 32B Preview
Experimental release of Qwen's reasoning model. Great at coding and math, but still in development so may exhibit odd bugs. Not production-ready.
Added Feb 27, 2025
Pricing
Auto routing · per 1M tokens- Input
- $0.20
- Output
- $0.20
Specifications
- Context window
- 32.8K
- Max output
- 32.8K
- Parameters
- 32B
Benchmarks
Benchmarks
Sourced from LMArena.
Arena Score
1154.2
Overall Rank
#376 / 419
Votes
3,231
Confidence Interval
1142.7 - 1165.7
Category Scores
Coding
#380 / 414
566 votes
1173.7
Math
#323 / 404
480 votes
1208.5
Longer Query
#360 / 397
475 votes
1179.0
Creative Writing
#382 / 417
454 votes
1103.5
Instruction Following
#370 / 419
1,329 votes
1144.9
Hard Prompts
#373 / 419
852 votes
1168.3
Additional Categories15
Chinese
#331 / 395
206 votes
1233.6
Industry Mathematical
#331 / 402
422 votes
1215.1
Expert
#352 / 369
232 votes
1154.1
Multi Turn
#361 / 417
487 votes
1145.4
Russian
#365 / 383
527 votes
1112.1
Non English
#366 / 419
1,448 votes
1125.3
Industry Legal And Government
#372 / 392
214 votes
1154.8
Exclude Ties
#374 / 419
2,230 votes
1011.2
Hard Prompts English
#375 / 417
519 votes
1177.1
English
#381 / 419
1,783 votes
1175.7
Industry Software And It Services
#381 / 419
824 votes
1166.7
Industry Life And Physical And Social Science
#384 / 417
583 votes
1159.5
Industry Business And Management And Financial Operations
#385 / 412
374 votes
1108.7
Industry Entertainment And Sports And Media
#385 / 417
540 votes
1102.1
Industry Writing And Literature And Language
#387 / 418
855 votes
1112.6
Published 2026-11-06 · Matched as QwQ-32B-Preview
LMArena Dataset