AlibabaQwen3 7
Qwen3 7 is an Alibaba proprietary reasoning model, released 19 May 2026. Its best configuration (max reasoning effort) ranks #106 of 775 on the BenchLeader Index at 60.7 ±3.9, in the upper half. The ± is the point: 200 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 instruction following (78) and lowest in agents & tools (49). At $3.75 per million tokens blended it is among the most expensive ranked models. Output speed of 201 tokens per second puts it in the fastest quarter, with a first answer in 14.3 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 19 Oct 2026.
- Blended price
- $3.75/M
- $2.50 in · $7.50 out
- Output speed
- 201 tok/s
- measured by Artificial Analysis
- First answer
- 14 s
- first token 1.07 s
- Context
- 1M
- Full answer
- 17 s
- median, reasoning included
- Cost per run
- $1.15
- one full Intelligence Index run
- Released
- 19 May 2026
- Overall index61
- Reasoning65
- Coding60
- Agents & tools49
- Maths59
- Knowledge63
- Instruction following78
- Human preference57
- Long context65
- Composite63
Versions
Alibaba has shipped 3 models under this name. Each is ranked on its own results; a newer version often has fewer results so far, which holds its index nearer the average until more arrive.
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 | – | – | 78 tok/s | 1.07 s | – | |
| xhigh | – | – | 78 tok/s | 1.07 s | – | |
| maxbest | 60.7 | #106 | 201 tok/s | 14 s | $0.0033 | Agents & tools 49 · Coding 60 · Composite 63 · Human preference 57 · Instruction following 78 · Knowledge 63 · Long context 65 · Maths 59 · Reasoning 65 |
How its index has moved
8 Sept 2026 to 19 Oct 2026The 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 | max | Source | Trend |
|---|---|---|---|---|
| GPQA Diamond | – | 90.9%#34 | Epoch AI Benchmarking Hub | |
| SimpleBench | – | 70.4%#13 | SimpleBench | |
| Arena Hard Prompts | – | 1495#42 | Arena (LMArena) | |
| LiveBench 1.0 Reasoningnot in index | – | 83.3%#43 | LiveBench | |
| GPQA Diamond (AA)not in index | – | 92.3%#36 | Artificial Analysis | |
| Humanity's Last Exam (AA)not in index | – | 40.5%#83 | Artificial Analysis | |
| GPQA Diamond (Vals)not in index | – | 90.2%#25 | Vals AI | |
| CritPt | – | 13.4%#100 | Artificial Analysis | |
| Chess Puzzlesnot in index | – | 26.0%#75 | Epoch AI Benchmarking Hub | |
| Mystery Game Puzzlesnot in index | 26.0%#42 | 32.0%#29 | Epoch AI Benchmarking Hub | |
| LMCAnot in index | – | 44.0%#96 | Epoch AI Benchmarking Hub | |
| DTBenchnot in index | – | 92.3%#70 | Epoch AI Benchmarking Hub |
Coding
| Benchmark | max | Source | Trend |
|---|---|---|---|
| SWE-bench Verified (Epoch) | 77.3%#7 | Epoch AI Benchmarking Hub | |
| SciCode | 49.5%#114 | SciCode | |
| Arena Coding | 1524#28 | Arena (LMArena) | |
| Arena WebDev | 1516#54 | Arena (LMArena) | |
| LiveBench 1.0 Codingnot in index | 74.2%#53 | LiveBench | |
| SciCode (AA)not in index | 49.5%#114 | Artificial Analysis | |
| LiveCodeBench | 87.1%#18 | Vals AI | |
| SWE-bench (Vals)not in index | 68.8%#66 | Vals AI | |
| Code Migrationnot in index | 13.1%#58 | Vals AI | |
| Vibe Code Bench v1.1not in index | 47.7%#62 | Vals AI | |
| ALE-Benchnot in index | 1189.4#43 | Epoch AI Benchmarking Hub | |
| Terminal-Bench 4.0 (AA)not in index | 1.5%#119 | Artificial Analysis | |
| Terminal-Bench 2.1 (AA)not in index | 74.5%#65 | Artificial Analysis | |
| Startrise Frontendnot in index | 63.6%#27 | Startrise |
Agents & tools
| Benchmark | max | Source | Trend |
|---|---|---|---|
| Arena Agent | -6.3#41 | Arena (LMArena) | |
| LiveBench Agentic Codingnot in index | 43.6%#56 | LiveBench | |
| Terminal-Bench Hard | 50.8%#22 | Artificial Analysis | |
| τ²-Bench Telecom (AA)not in index | 94.7%#29 | Artificial Analysis | |
| Terminal-Bench 2.1 (Vals) | 61.0%#39 | Vals AI | |
| GDPval-AA v2.1not in index | 31.4%#131 | Artificial Analysis | |
| τ³-Banking (AA) | 11.8%#134 | Artificial Analysis | |
| ITBench SRE (AA)not in index | 42.5%#15 | Artificial Analysis | |
| Analyst Agent (AA)not in index | 18.8%#37 | Artificial Analysis | |
| Finance Agent v2not in index | 47.8%#51 | Vals AI | |
| Legal Research Benchnot in index | 25.5%#53 | Vals AI | |
| Tax Agent Benchnot in index | 15.8%#49 | Vals AI | |
| Terminal-Bench 2.0 (Vals)not in index | 59.2%#9 | Vals AI | |
| GBAEvalnot in index | 0.4%#20 | Epoch AI Benchmarking Hub | |
| AutomationBenchnot in index | 23.0%#127 | Artificial Analysis | |
| GDP.pdf (AA)not in index | 8.6%#137 | Artificial Analysis | |
| EnterpriseOps-Gymnot in index | 45.0%#14 | Artificial Analysis |
Maths
| Benchmark | max | Source | Trend |
|---|---|---|---|
| FrontierMath Tiers 1–3 | 64.6%#36 | Epoch AI Benchmarking Hub | |
| FrontierMath Tier 4 | 34.1%#34 | Epoch AI Benchmarking Hub | |
| OTIS Mock AIME | 95.6%#43 | Epoch AI Benchmarking Hub | |
| ProofBench | 26.0%#44 | Vals AI | |
| LiveBench Mathematicsnot in index | 85.3%#54 | LiveBench | |
| Arena Maths | 1486#32 | Arena (LMArena) |
Knowledge
| Benchmark | xhigh | max | Source | Trend |
|---|---|---|---|---|
| SimpleQA Verified | – | 55.8%#21 | Epoch AI Benchmarking Hub | |
| LiveBench 1.0 Data Analysisnot in index | – | 71.8%#46 | LiveBench | |
| AA-Omniscience | – | 13.5#88 | Artificial Analysis | |
| MMLU-Pro | – | 89.3%#12 | Vals AI | |
| LegalBench | – | 84.9%#28 | Vals AI | |
| CorpFin | – | 63.7%#41 | Vals AI | |
| TaxEval | – | 75.3%#19 | Vals AI | |
| Excel Modeling Benchmarknot in index | – | 57.0%#41 | Vals AI | |
| MedCodenot in index | – | 38.8%#67 | Vals AI | |
| MedScribenot in index | – | 79.4%#61 | Vals AI | |
| DrugDiscoveryBenchnot in index | 29.3%#15 | – | Scale AI SEAL | |
| EBR-benchnot in index | – | 9.5%#22 | Epoch AI Benchmarking Hub | |
| AA-Omniscience: accuracynot in index | – | 31.1%#171 | Artificial Analysis | |
| AA-Omniscience: non-hallucinationnot in index | – | 74.4%#25 | Artificial Analysis |
Instruction following
| Benchmark | max | Source | Trend |
|---|---|---|---|
| LiveBench Languagenot in index | 79.7%#33 | LiveBench | |
| LiveBench Instruction Followingnot in index | 74.0%#14 | LiveBench | |
| IFBench | 80.5%#7 | Artificial Analysis | |
| Arena Instruction Followingnot in index | 1465#44 | Arena (LMArena) |
Human preference
| Benchmark | max | Source | Trend |
|---|---|---|---|
| Arena Text | 1474#39 | Arena (LMArena) | |
| EQ-Bench 4 | 1110#22 | EQ-Bench | |
| Arena Creative Writingnot in index | 1439#59 | Arena (LMArena) | |
| Arena Multi-turnnot in index | 1483#35 | Arena (LMArena) | |
| Arena Longer Queriesnot in index | 1491#28 | Arena (LMArena) |
Long context
| Benchmark | max | Source | Trend |
|---|---|---|---|
| AA-LCR | 79.0%#105 | Artificial Analysis |
Composite
| Benchmark | max | Source | Trend |
|---|---|---|---|
| Epoch Capabilities Indexnot in index | 154.0#44 | Epoch AI Benchmarking Hub | |
| LiveBench 1.0not in index | 73.1%#47 | LiveBench | |
| AA Intelligence Index v4.3.2 | 29.5#126 | 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 |
|---|---|---|---|---|---|---|
| Alibaba Cloud Int. | 85 tok/s | 1.43 s | $2.00 | $6.00 | 1M | – |
| Alibaba Cloud Int. | 82 tok/s | 1.07 s | $1.48 | $4.42 | 1M | – |
Price history
Listed price per 1M tokens over time, as recorded by OpenRouter for the provider with the longest history.
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 at $0.330 per 1M. Reasoning tokens are not modelled.
| Workload | Tokens in / out | Cost | With caching | Time |
|---|---|---|---|---|
| Chat reply | 400 / 300 | $0.0033 | $0.0026 | 15.8 s |
| Summarize a 30-page report | 12,000 / 600 | $0.035 | $0.015 | 17.3 s |
| Code edit | 6,000 / 1,500 | $0.026 | $0.017 | 21.7 s |
| Agentic coding session | 60,000 / 4,000 | $0.180 | $0.082 | 34.2 s |
| Structured extraction | 2,000 / 200 | $0.0065 | $0.0032 | 15.3 s |
More from Alibaba
Compare Qwen3 7 with…
See also
Data as of 21 Oct 2026. Compare these configurations.
Cite as: BenchLeader, “Qwen3 7: benchmarks, pricing, speed and rank”, https://www.benchleader.com/models/qwen3-7, data as of 21 Oct 2026.