AlibabaReasoning modelQwen3.6 Max
Qwen3.6 Max is an Alibaba proprietary reasoning model, released 20 Apr 2026. Its best configuration (max reasoning effort) ranks #80 of 756 on the BenchLeader Index at 62.5 ±2.9, in the upper half. The ± is the point: 155 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 (74) and lowest in coding (57). At $2.92 per million tokens blended it is pricier than most ranked models. Output speed of 73 tokens per second puts it slower than most, with a first answer in 30.4 s. It has been measured at 2 reasoning-effort settings; this summary describes the best-scoring one, and the tabs above switch between them. Last measured 2 Oct 2026.
- Blended price
- $2.92/M
- $1.30 in · $7.80 out
- Output speed
- 73 tok/s
- measured by Artificial Analysis
- First answer
- 30 s
- first token 1.00 s
- Context
- 262k
- Full answer
- 37 s
- median, reasoning included
- Cost per run
- –
- Released
- 20 Apr 2026
- Overall index63
- Reasoning59
- Coding57
- Agents & tools72
- Maths66
- Knowledge63
- Instruction following74
- Human preference65
- Long context66
- Composite62
Versions
Alibaba has shipped 2 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.
| Model | Released | Index | Rank |
|---|---|---|---|
| Qwen3.8 Maxmax | 3 Aug 2026 | 65.0 | #43 |
| Qwen3.6 Maxmaxthis page | 20 Apr 2026 | 62.5 | #80 |
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 | – | – | 64 tok/s | 1.00 s | – | |
| maxbest | 62.5 | #80 | 73 tok/s | 30 s | $0.0029 | Agents & tools 72 · Coding 57 · Composite 62 · Human preference 65 · Instruction following 74 · Knowledge 63 · Long context 66 · Maths 66 · Reasoning 59 |
How its index has moved
9 Sept 2026 to 3 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
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 | – | 87.4%#67 | Epoch AI Benchmarking Hub | |
| SimpleBench | – | 63.0%#20 | SimpleBench | |
| LMArena Hard Prompts | – | 1482#63 | LMArena | |
| GPQA Diamond (AA)not in index | – | 88.8%#83 | Artificial Analysis | |
| Humanity's Last Exam (AA)not in index | – | 30.8%#127 | Artificial Analysis | |
| CritPt | – | 3.7%#152 | Artificial Analysis | |
| Chess Puzzlesnot in index | – | 20.0%#85 | Epoch AI Benchmarking Hub | |
| Mystery Game Puzzlesnot in index | 19.0%#59 | – | Epoch AI Benchmarking Hub | |
| LMCAnot in index | – | 42.5%#82 | Epoch AI Benchmarking Hub | |
| DTBenchnot in index | – | 87.2%#85 | Epoch AI Benchmarking Hub |
Coding
| Benchmark | no reasoning | max | Source | Trend |
|---|---|---|---|---|
| SWE-bench Verified (Epoch) | – | 76.7%#9 | Epoch AI Benchmarking Hub | |
| LMArena Coding | – | 1512#53 | LMArena | |
| LMArena WebDev | – | 1482#57 | LMArena | |
| SWE-bench (Vals)not in index | – | 72.8%#52 | Vals AI |
Agents & tools
| Benchmark | no reasoning | max | Source | Trend |
|---|---|---|---|---|
| Terminal-Bench Hard | – | 43.9%#37 | Artificial Analysis | |
| τ²-Bench Telecom (AA)not in index | – | 95.9%#16 | Artificial Analysis | |
| Terminal-Bench 2.0 (Vals)not in index | – | 51.7%#18 | Vals AI | |
| Vending-Bench 2not in index | – | 4254.2#38 | Andon Labs |
Maths
| Benchmark | no reasoning | max | Source | Trend |
|---|---|---|---|---|
| OTIS Mock AIME | – | 91.1%#62 | Epoch AI Benchmarking Hub | |
| LMArena Maths | – | 1476#40 | LMArena |
Knowledge
| Benchmark | no reasoning | max | Source | Trend |
|---|---|---|---|---|
| SimpleQA Verified | – | 52.0%#22 | Epoch AI Benchmarking Hub | |
| AA-Omniscience | – | 9.2#91 | Artificial Analysis | |
| CorpFin | – | 66.5%#18 | Vals AI | |
| CaseLaw v2not in index | – | 47.9%#50 | Vals AI | |
| AA-Omniscience: accuracynot in index | – | 37.9%#131 | Artificial Analysis | |
| AA-Omniscience: non-hallucinationnot in index | – | 53.8%#89 | Artificial Analysis |
Instruction following
| Benchmark | no reasoning | max | Source | Trend |
|---|---|---|---|---|
| IFBench | – | 76.6%#19 | Artificial Analysis | |
| LMArena Instruction Followingnot in index | – | 1450#65 | LMArena |
Human preference
| Benchmark | no reasoning | max | Source | Trend |
|---|---|---|---|---|
| LMArena Text | – | 1460#59 | LMArena | |
| LMArena Creative Writingnot in index | – | 1437#60 | LMArena | |
| LMArena Multi-turnnot in index | – | 1469#64 | LMArena | |
| LMArena Longer Queriesnot in index | – | 1474#55 | LMArena |
Long context
| Benchmark | no reasoning | max | Source | Trend |
|---|---|---|---|---|
| AA-LCR | – | 80.7%#58 | Artificial Analysis |
Composite
| Benchmark | no reasoning | max | Source | Trend |
|---|---|---|---|---|
| Epoch Capabilities Indexnot in index | – | 149.2#54 | Epoch AI Benchmarking Hub | |
| AA Intelligence Index | – | 28.4#119 | 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. | 66 tok/s | 1.41 s | $1.03 | $6.16 | 262k | – |
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 $0.130 per 1M. Reasoning tokens are not modelled.
| Workload | Tokens in / out | Cost | With caching | Time |
|---|---|---|---|---|
| Chat reply | 400 / 300 | $0.0029 | $0.0025 | 34.5 s |
| Summarise a 30-page report | 12,000 / 600 | $0.020 | $0.0098 | 38.6 s |
| Code edit | 6,000 / 1,500 | $0.019 | $0.014 | 50.9 s |
| Agentic coding session | 60,000 / 4,000 | $0.109 | $0.057 | 1.4 min |
| Structured extraction | 2,000 / 200 | $0.0042 | $0.0024 | 33.1 s |
See also
Data as of 5 Oct 2026. Compare these configurations.
Cite as: BenchLeader, “Qwen3.6 Max: benchmarks, pricing, speed and rank”, https://www.benchleader.com/models/qwen3-6-max, data as of 5 Oct 2026.