AlibabaQwen3 Max
Qwen3 Max is an Alibaba proprietary model, released 5 Sept 2025. Its best configuration (max reasoning effort) ranks #300 of 785 on the BenchLeader Index at 52.8 ±3.1, in the upper half. The ± is the point: 286 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 human preference (62) and lowest in composite (47). At $0.628 per million tokens blended it is mid-priced. Last measured 22 Oct 2026.
- Blended price
- $0.628/M
- $0.359 in · $1.43 out
- Output speed
- –
- First answer
- –
- Context
- 262k
- Full answer
- –
- Cost per run
- –
- Released
- 5 Sept 2025
- Overall index53
- Reasoning53
- Coding58
- Agents & tools52
- Maths54
- Knowledge54
- Instruction following50
- Human preference62
- Long context49
- Composite47
How its index has moved
8 Sept 2026 to 21 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 | max | Source | Trend |
|---|---|---|---|
| GPQA Diamond | 72.6%#161 | Epoch AI Benchmarking Hub | |
| Arena Hard Prompts | 1458#99 | Arena (LMArena) | |
| GPQA Diamond (AA)not in index | 76.4%#235 | Artificial Analysis | |
| Humanity's Last Exam (AA)not in index | 11.9%#287 | Artificial Analysis | |
| GPQA Diamond (Vals)not in index | 84.8%#48 | Vals AI | |
| Kagi LLM Benchmark | 72.5%#22 | Kagi LLM Benchmark | |
| CritPt | 0.9%#237 | Artificial Analysis | |
| Mystery Game Puzzlesnot in index | 5.0%#132 | Epoch AI Benchmarking Hub | |
| CL-benchnot in index | 14.5%#19 | Epoch AI Benchmarking Hub | |
| LMCAnot in index | 28.3%#144 | Epoch AI Benchmarking Hub | |
| DTBenchnot in index | 82.1%#122 | Epoch AI Benchmarking Hub |
Coding
| Benchmark | max | Source | Trend |
|---|---|---|---|
| Arena Coding | 1480#105 | Arena (LMArena) | |
| LiveCodeBench | 78.2%#74 | Vals AI | |
| Vibe Code Bench v1.1not in index | 3.5%#105 | Vals AI | |
| ALE-Benchnot in index | 370.4#110 | Epoch AI Benchmarking Hub |
Agents & tools
| Benchmark | max | Source | Trend |
|---|---|---|---|
| Terminal-Bench Hard | 20.4%#163 | Artificial Analysis | |
| τ²-Bench Telecom (AA)not in index | 74.3%#141 | Artificial Analysis | |
| Poker Agentnot in index | 994.5%#14 | Vals AI | |
| Terminal-Bench 2.0 (Vals)not in index | 24.7%#51 | Vals AI | |
| Vending-Bench 2not in index | 71.6#63 | Andon Labs |
Maths
| Benchmark | max | Source | Trend |
|---|---|---|---|
| FrontierMath Tiers 1–3 | 18.9%#93 | Epoch AI Benchmarking Hub | |
| OTIS Mock AIME | 73.3%#129 | Epoch AI Benchmarking Hub | |
| MATH Level 5 | 97.1%#7 | Epoch AI Benchmarking Hub | |
| AIME (Vals) | 81.0%#46 | Vals AI | |
| MGSM | 92.2%#21 | Vals AI | |
| Arena Maths | 1433#102 | Arena (LMArena) |
Knowledge
| Benchmark | max | Source | Trend |
|---|---|---|---|
| SimpleQA Verified | 48.8%#31 | Epoch AI Benchmarking Hub | |
| AA-Omniscience | -42.8#339 | Artificial Analysis | |
| MMLU-Pro | 85.0%#58 | Vals AI | |
| LegalBench | 81.9%#78 | Vals AI | |
| CorpFin | 68.0%#11 | Vals AI | |
| TaxEval | 74.0%#42 | Vals AI | |
| MedQA | 87.4%#55 | Vals AI | |
| CaseLaw v2not in index | 55.0%#35 | Vals AI | |
| MedCodenot in index | 31.4%#90 | Vals AI | |
| MedScribenot in index | 72.7%#84 | Vals AI | |
| AA-Omniscience: accuracynot in index | 24.4%#243 | Artificial Analysis | |
| AA-Omniscience: non-hallucinationnot in index | 11.6%#388 | Artificial Analysis |
Instruction following
| Benchmark | max | Source | Trend |
|---|---|---|---|
| IFBench | 48.0%#186 | Artificial Analysis | |
| Arena Instruction Followingnot in index | 1424#105 | Arena (LMArena) |
Human preference
| Benchmark | max | Source | Trend |
|---|---|---|---|
| Arena Text | 1434#104 | Arena (LMArena) | |
| Arena Creative Writingnot in index | 1393#114 | Arena (LMArena) | |
| Arena Multi-turnnot in index | 1447#92 | Arena (LMArena) | |
| Arena Longer Queriesnot in index | 1443#102 | Arena (LMArena) |
Long context
| Benchmark | max | Source | Trend |
|---|---|---|---|
| AA-LCR | 50.0%#321 | Artificial Analysis |
Composite
| Benchmark | max | Source | Trend |
|---|---|---|---|
| Epoch Capabilities Indexnot in index | 142.3#102 | Epoch AI Benchmarking Hub | |
| AA Intelligence Index v4.3.2 | 15.6#277 | Artificial Analysis |
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 at $0.240 per 1M. Reasoning tokens are not modelled.
| Workload | Tokens in / out | Cost | With caching | Time |
|---|---|---|---|---|
| Chat reply | 400 / 300 | $0.0006 | $0.0005 | – |
| Summarize a 30-page report | 12,000 / 600 | $0.0052 | $0.0041 | – |
| Code edit | 6,000 / 1,500 | $0.0043 | $0.0038 | – |
| Agentic coding session | 60,000 / 4,000 | $0.027 | $0.022 | – |
| Structured extraction | 2,000 / 200 | $0.0010 | $0.0008 | – |
More from Alibaba
See also
Questions people ask
- What are Qwen3 Max's benchmark scores?
- As of 25 Oct 2026, Qwen3 Max (max effort) scores 52.8 ±3.1 on the BenchLeader Index, strongest in human preference (62) and weakest in composite (47). That rests on 24 independent benchmark results, and every score on this page links to its publisher.
- How does Qwen3 Max rank?
- Qwen3 Max (max effort) is #300 of 785 ranked configurations on 25 Oct 2026. 286 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 Max cost per 1M tokens?
- As of 25 Oct 2026, Qwen3 Max costs $0.359 per 1M input tokens and $1.43 per 1M output tokens, $0.628 blended (three input tokens per output token). The list price last changed on 13 Oct 2026.
- When was Qwen3 Max released?
- Alibaba released Qwen3 Max on 5 Sept 2025.
Data as of 25 Oct 2026. Compare with another model.
Cite as: BenchLeader, “Qwen3 Max: benchmarks, pricing, speed and rank”, data as of 25 Oct 2026, https://www.benchleader.com/models/qwen3-max.