LLMs TigerLive LLM benchmarks

QwQ-32B: benchmark scores, price and ranking

QwQ-32B from Alibaba ranks #237 of 743 LLMs on the LLMs Tiger aggregate benchmark score (49.7), based on 6 benchmark results captured up to 2026-10-13.

Organisation
Alibaba
Type
Open-weight / local model
Open weights
Yes
Released
2025-03-06 (source: curated)
Cheapest API price
$0.150 in / $0.400 out per million tokens (via deepinfra, observed 2026-10-13T20:47Z)
Aggregate rank
#237 of 743 (score 49.7)

Relative to other models, QwQ-32B is strongest on Fiction.liveBench (#6 of 47) and weakest on Aider Polyglot (#37 of 49).

Across the 6 benchmarks where at least ten models are scored, QwQ-32B is in the top 10% on 0 and in the top half on 3.

Its cheapest listed input price, $0.150 per million tokens, is lower than that of 72% of the 427 models with a listed price.

QwQ-32B benchmark results

QwQ-32B benchmark scores, ranked against all models measured on each benchmark
BenchmarkScoreRankSourceCaptured
Fiction.liveBench83.3%#6 of 47Epoch AI Benchmarking Hub2026-10-13
Epoch ECI138#61 of 204Epoch AI Benchmarking Hub2026-10-13
OTIS Mock AIME59.2%#93 of 189Epoch AI Benchmarking Hub2026-10-13
GPQA Diamond65.3%#159 of 275Epoch AI Benchmarking Hub2026-10-13
Chess Puzzles5.0%#82 of 122Epoch AI Benchmarking Hub2026-10-13
Aider Polyglot20.9%#37 of 49Aider polyglot leaderboard2026-10-13

About QwQ-32B

QwQ-32B is an open-weight model that can run on your own hardware from Alibaba. It was released on 2025-03-06, according to the developer's announcement. The cheapest API listing we track charges $0.150 per million input tokens and $0.400 per million output tokens, through deepinfra. We hold 6 benchmark results for it across 3 areas, and none is in the top tenth of its board.

Where QwQ-32B stands

On our aggregate score, QwQ-32B is #237 of 743 ranked models, around the middle of the table. Its score is 49.7; the scale is centred on 50, so it sits 0.3 points below the typical model. That beats 68% of the ranked field. Among the 74 ranked Alibaba models it is #18. The next one up is Qwen2.5-32B at 49.7. The next one down is Qwen3-235B-A22B-Thinking-2507 at 49.7.

QwQ-32B, benchmark by benchmark

Fiction.liveBench puts QwQ-32B at 83.3%, #6 of 47. That beats 87% of the models tested, but 11.1 points short of grok-4-0709 (94.4%). The results either side are DeepSeek-V3.2-Exp (high) (83.3%) and gemini-2.5-flash-preview-05-20 (77.8%). It covers long-context story comprehension (16k). The board median is 61.1%, so QwQ-32B is 22.2 points above it. Source: Epoch AI Benchmarking Hub, captured 2026-10-13.

On Epoch ECI, QwQ-32B reaches 138. It ranks #61 of 204, ahead of 70% of the field, and 19 points behind Qwen 3.8 Max at 156. The results either side are qwen3-14b (138) and DeepSeek-R1-Distill-Qwen-32B (137). It covers Epoch Capabilities Index: Epoch's own aggregate of ~40 benchmarks. The board median is 127, so QwQ-32B is 10 points above it. Source: Epoch AI Benchmarking Hub, captured 2026-10-13.

QwQ-32B manages 59.2% on OTIS Mock AIME, good for #93 of 189. Claude Fable 5.1 (max) leads at 100.0%, 40.8 points further up, and 51% of models sit below. The results either side are Qwen3.5 9B (61.7%) and glm-4.7-flash (58.3%). It covers competition maths (AIME 2024-25 style). The board median is 57.8%, so QwQ-32B is 1.4 points above it. Source: Epoch AI Benchmarking Hub, captured 2026-10-13.

QwQ-32B lands below the middle on GPQA Diamond: 65.3%, #159 of 275, with 57% of the field scoring higher. GPT-6 Astra (max) leads at 95.8%; the closest result above is qwq-plus at 65.4%. It covers expert-level science questions. The board median is 71.7%, so QwQ-32B is 6.4 points below it. Source: Epoch AI Benchmarking Hub, captured 2026-10-13.

QwQ-32B lands below the middle on Chess Puzzles: 5.0%, #82 of 122, with 66% of the field scoring higher. GPT-6 Astra (max) leads at 72.0%; the closest result above is qwen3-8b at 5.0%. It covers chess tactics puzzles. The board median is 12.5%, so QwQ-32B is 7.5 points below it. Source: Epoch AI Benchmarking Hub, captured 2026-10-13.

QwQ-32B lands below the middle on Aider Polyglot: 20.9%, #37 of 49, with 73% of the field scoring higher. gemini-2.5-pro-preview-06-05 leads at 83.1%; the closest result above is qwen-max-2025-01-25 at 21.8%. It covers edits across 6 languages (225 exercises). The board median is 44.9%, so QwQ-32B is 24.0 points below it. Source: Aider polyglot leaderboard, captured 2026-10-13.

Strengths and weaknesses

Reasoning, maths and science: uneven, ahead of 46% of the field on average across 2 benchmarks. Its best board here is OTIS Mock AIME (#93 of 189) and its weakest is GPQA Diamond (#159 of 275). General knowledge and chat: solid, ahead of 63% of the field on average across 3 benchmarks. Its best board here is Fiction.liveBench (#6 of 47) and its weakest is Chess Puzzles (#82 of 122).

Price and value

The median input price across the 427 priced models is $0.500 per million tokens, so QwQ-32B costs less than the typical listing. Models that score within 1.5 points or better and cost less: Qwen2.5-Coder-7B-Instruct ($0.010, 49.5), DeepSeek V4.1 Flash (high) ($0.010, 48.9), DeepSeek V4 Flash 0731 (max) ($0.016, 53.1). Similar scores cost more on Qwen3.5 397B-A17B ($0.164) and Phi-3-medium-4k-instruct ($0.170).

Release history

The Alibaba model released before it was qwen-plus-2025-01-25 on 2025-01-25, which scores 45.7. The next one was qwq-plus on 2025-04-08, at 49.3.

Reading these numbers

The 6 scores on this page come from 2 sources: Epoch AI Benchmarking Hub (5), Aider polyglot leaderboard (1). 1 of them come from an official leaderboard and 5 from datasets or third-party runs. With 6 results the aggregate rank can move a lot when one new score arrives.

Models ranked nearby

More from Alibaba

Compare it on the live leaderboard → · How the aggregate score works