AlibabaQwen3 32B
Qwen3 32B is an Alibaba open-weights reasoning model, released 29 Apr 2025. Its best configuration ranks #344 of 785 on the BenchLeader Index at 51.5 ±4.6, in the upper half. The ± is the point: 312 other configurations score within that range, so they and this one cannot be told apart on quality alone — price and speed are what separate them. It scores highest in agents & tools (59) and lowest in coding (45). At $1.23 per million tokens blended it is pricier than most ranked models. Output speed of 29 tokens per second puts it in the slowest quarter, with a first token in 0.4 s. It has been measured at 3 reasoning-effort settings; this summary describes the best-scoring one, and the tabs above switch between them. Last measured 22 Oct 2026.
- Blended price
- $1.23/M
- $0.700 in · $2.80 out
- Output speed
- 29 tok/s
- OpenRouter traffic, 7-day median; not yet measured by Artificial Analysis
- First answer
- 0.38 s
- Context
- 131k
- Full answer
- –
- Cost per run
- –
- Released
- 29 Apr 2025
- Overall index52
- Reasoning50
- Coding45
- Agents & tools59
- Maths56
- Human preference51
Reasoning-effort configurations
The same model behaves differently depending on how much it is allowed to think. Each row is one setting, scored only on the benchmarks that were run at that setting. “Not stated” collects results from publishers that did not say which setting they used; for a reasoning model that is usually its thinking mode, but we do not assume it. Pick a setting here or at the top of the page to see its price, speed and category scores.
| Effort | Index | Rank | Speed | First answer | Chat reply cost | Categories |
|---|---|---|---|---|---|---|
| no reasoning | 39.6 | #682 | 58 tok/s | 1.38 s | $0.0002 | Composite 38 · Instruction following 35 · Long context 23 · Maths 34 · Reasoning 42 |
| thinking | 40.0 | #666 | 23 tok/s | 87 s | $0.0003 | Agents & tools 34 · Composite 38 · Instruction following 40 · Knowledge 39 · Long context 23 · Reasoning 41 |
| not statedbest | 51.5 | #344 | 29 tok/s | 0.38 s | $0.0011 | Agents & tools 59 · Coding 45 · Human preference 51 · Maths 56 · Reasoning 50 |
How its index has moved
8 Sept 2026 to 23 Oct 2026- not stated
- thinking
- no reasoning
The index is recomputed from scratch every day, so a line moves when a new benchmark result lands, when a publisher revises a score, or when the models it is normalised against change. Early movement usually means the score is still settling.
Benchmark results
See all LLM benchmarks. One column per reasoning effort. Rank is among every configuration of every model on that benchmark. Hover a score for the run it came from.
Reasoning
| Benchmark | no reasoning | thinking | not stated | Source | Trend |
|---|---|---|---|---|---|
| GPQA Diamond | 54.1%#206 | – | 65.7%#180 | Epoch AI Benchmarking Hub | |
| Arena Hard Promptseffort not stated | – | – | 1367#199 | Arena (LMArena) | |
| GPQA Diamond (AA)not in index | 53.5%#411 | 66.8%#328 | – | Artificial Analysis | |
| Humanity's Last Exam (AA)not in index | 4.1%#536 | 7.4%#365 | – | Artificial Analysis | |
| Kagi LLM Benchmark | 48.7%#88 | – | – | Kagi LLM Benchmark | |
| CritPt | – | 0.3%#271 | – | Artificial Analysis | |
| SciPredictnot in indexeffort not stated | – | – | 17.0%#11 | Scale AI SEAL | |
| LMCAnot in indexeffort not stated | – | – | 17.3%#165 | Epoch AI Benchmarking Hub | |
| DTBenchnot in indexeffort not stated | – | – | 67.5%#164 | Epoch AI Benchmarking Hub |
Coding
| Benchmark | thinking | not stated | Source | Trend |
|---|---|---|---|---|
| SciCodeeffort not stated | – | 35.4%#191 | SciCode | |
| Arena Codingeffort not stated | – | 1406#190 | Arena (LMArena) | |
| SciCode (AA)not in index | 36.0%#199 | – | Artificial Analysis | |
| Aider Polygloteffort not stated | – | 40.0%#31 | Aider polyglot leaderboard | |
| Terminal-Bench 4.0 (AA)not in index | 0.0%#152 | – | Artificial Analysis | |
| Terminal-Bench 2.1 (AA)not in index | 5.2%#205 | – | Artificial Analysis |
Agents & tools
| Benchmark | thinking | not stated | Source | Trend |
|---|---|---|---|---|
| Terminal-Bench Hard | 3.0%#314 | – | Artificial Analysis | |
| τ²-Bench Telecom (AA)not in index | 29.8%#260 | – | Artificial Analysis | |
| BFCL Overalleffort not stated | – | 48.7%#25 | Berkeley Function Calling Leaderboard | |
| τ³-Banking (AA) | 5.4%#179 | – | Artificial Analysis | |
| AutomationBenchnot in index | 1.0%#195 | – | Artificial Analysis | |
| GDP.pdf (AA)not in index | 1.6%#194 | – | Artificial Analysis | |
| BrowseComp-Plusnot in indexeffort not stated | – | 3.6%#21 | BrowseComp-Plus |
Maths
| Benchmark | no reasoning | not stated | Source | Trend |
|---|---|---|---|---|
| OTIS Mock AIME | 23.1%#217 | 66.9%#148 | Epoch AI Benchmarking Hub | |
| Arena Mathseffort not stated | – | 1399#151 | Arena (LMArena) |
Knowledge
| Benchmark | thinking | Source | Trend |
|---|---|---|---|
| AA-Omniscience | -50.4#389 | Artificial Analysis | |
| AA-Omniscience: accuracynot in index | 17.4%#365 | Artificial Analysis | |
| AA-Omniscience: non-hallucinationnot in index | 17.9%#306 | Artificial Analysis |
Instruction following
| Benchmark | no reasoning | thinking | not stated | Source | Trend |
|---|---|---|---|---|---|
| IFBench | 31.5%#351 | 36.3%#302 | – | Artificial Analysis | |
| Arena Instruction Followingnot in indexeffort not stated | – | – | 1331#208 | Arena (LMArena) |
Human preference
| Benchmark | not stated | Source | Trend |
|---|---|---|---|
| Arena Text | 1347#205 | Arena (LMArena) | |
| Arena Creative Writingnot in index | 1304#213 | Arena (LMArena) | |
| Arena Multi-turnnot in index | 1337#208 | Arena (LMArena) | |
| Arena Longer Queriesnot in index | 1355#200 | Arena (LMArena) |
Long context
| Benchmark | no reasoning | thinking | Source | Trend |
|---|---|---|---|---|
| AA-LCR | 0.0%#493 | 0.0%#493 | Artificial Analysis |
Composite
| Benchmark | no reasoning | thinking | not stated | Source | Trend |
|---|---|---|---|---|---|
| Epoch Capabilities Indexnot in indexeffort not stated | – | – | 138.4#120 | Epoch AI Benchmarking Hub | |
| AA Intelligence Index v4.3.2 | 7.3#480 | 7.2#488 | – | Artificial Analysis |
Where to run it
Every provider serving this model through OpenRouter, with throughput and first-token latency measured on live traffic over the last 30 minutes and each provider’s own price. Purple marks the best in each column.
| Provider | Speed | First token | Input $/M | Output $/M | Context | Quantisation |
|---|---|---|---|---|---|---|
| Tenstorrent | 44 tok/s | 0.50 s | $0.140 | $0.400 | 20k | fp8 |
| DeepInfra | 42 tok/s | 0.38 s | $0.080 | $0.280 | 41k | fp8 |
| SiliconFlow | 27 tok/s | 1.09 s | $0.140 | $0.570 | 131k | fp8 |
Price history
The price above, per 1M tokens, every day it changed.
What a task costs
Estimates from list price, output speed and time to first answer for the best configuration. “With caching” assumes three-quarters of the input is served from the prompt cache. Reasoning tokens are not modelled.
| Workload | Tokens in / out | Cost | With caching | Time |
|---|---|---|---|---|
| Chat reply | 400 / 300 | $0.0011 | – | 10.7 s |
| Summarize a 30-page report | 12,000 / 600 | $0.010 | – | 21.1 s |
| Code edit | 6,000 / 1,500 | $0.0084 | – | 52.1 s |
| Agentic coding session | 60,000 / 4,000 | $0.053 | – | 2.3 min |
| Structured extraction | 2,000 / 200 | $0.0020 | – | 7.3 s |
More from Alibaba
See also
Questions people ask
- What are Qwen3 32B's benchmark scores?
- As of 24 Oct 2026, Qwen3 32B scores 51.5 ±4.6 on the BenchLeader Index, strongest in agents & tools (59) and weakest in coding (45). That rests on 9 independent benchmark results, and every score on this page links to its publisher.
- How does Qwen3 32B rank?
- Qwen3 32B is #344 of 785 ranked configurations on 24 Oct 2026. 312 other configurations score within its margin, so the exact position matters less than the band. The leader is Claude Opus 5.5 (max effort) at 72.6.
- How much does Qwen3 32B cost per 1M tokens?
- As of 24 Oct 2026, Qwen3 32B costs $0.700 per 1M input tokens and $2.80 per 1M output tokens, $1.23 blended (three input tokens per output token). The list price last changed on 24 Oct 2026.
- When was Qwen3 32B released?
- Alibaba released Qwen3 32B on 29 Apr 2025. BenchLeader has results for it at 3 reasoning-effort settings.
Data as of 24 Oct 2026. Compare these configurations.
Cite as: BenchLeader, “Qwen3 32B: benchmarks, pricing, speed and rank”, data as of 24 Oct 2026, https://www.benchleader.com/models/qwen3-32b.