AlibabaQwen3.5-122B-A10B
Qwen3.5-122B-A10B is an Alibaba open-weights reasoning model, released 24 Feb 2026. Its best configuration ranks #256 of 785 on the BenchLeader Index at 54.3 ±5.9, in the upper half. The ± is the point: 333 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 maths (60) and lowest in coding (46). At $1.10 per million tokens blended it is pricier than most ranked models. Output speed of 57 tokens per second puts it slower than most, with a first token in 1.3 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 23 Oct 2026.
- Blended price
- $1.10/M
- $0.400 in · $3.20 out
- Output speed
- 57 tok/s
- OpenRouter traffic, 7-day median; not yet measured by Artificial Analysis
- First answer
- 1.29 s
- Context
- 262k
- Full answer
- –
- Cost per run
- –
- Released
- 24 Feb 2026
- Overall index54
- Reasoning59
- Coding46
- Maths60
- Human preference59
- Multimodal57
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 | 47.0 | #465 | 142 tok/s | 2.25 s | $0.0011 | Agents & tools 45 · Coding 34 · Composite 49 · Instruction following 52 · Knowledge 37 · Long context 55 · Multimodal 53 · Reasoning 42 |
| thinking | 52.6 | #311 | 131 tok/s | 18 s | $0.0011 | Agents & tools 49 · Composite 47 · Instruction following 74 · Knowledge 43 · Long context 63 · Multimodal 58 · Reasoning 42 |
| not statedbest | 54.3 | #256 | 57 tok/s | 1.29 s | $0.0011 | Coding 46 · Human preference 59 · Maths 60 · Multimodal 57 · Reasoning 59 |
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.
Metadata
| Benchmark | thinking | Source | Trend |
|---|---|---|---|
| AA Openness Indexnot in index | 38.9#34 | Artificial Analysis |
Reasoning
| Benchmark | no reasoning | thinking | not stated | Source | Trend |
|---|---|---|---|---|---|
| Arena Hard Promptseffort not stated | – | – | 1433#129 | Arena (LMArena) | |
| GPQA Diamond (AA)not in index | 82.7%#168 | 85.7%#120 | – | Artificial Analysis | |
| Humanity's Last Exam (AA)not in index | 15.9%#245 | 25.2%#184 | – | Artificial Analysis | |
| CritPt | 0.9%#237 | 0.6%#254 | – | Artificial Analysis | |
| Mystery Game Puzzlesnot in index | 17.0%#75 | – | – | Epoch AI Benchmarking Hub | |
| LMCAnot in index | 32.2%#132 | – | – | Epoch AI Benchmarking Hub | |
| DTBenchnot in index | 84.3%#109 | – | – | Epoch AI Benchmarking Hub |
Coding
| Benchmark | no reasoning | thinking | not stated | Source | Trend |
|---|---|---|---|---|---|
| SciCode | 35.6%#189 | – | – | SciCode | |
| Arena Codingeffort not stated | – | – | 1459#133 | Arena (LMArena) | |
| Arena WebDeveffort not stated | – | – | 1361#105 | Arena (LMArena) | |
| SciCode (AA)not in index | – | 39.7%#178 | – | Artificial Analysis | |
| Terminal-Bench 4.0 (AA)not in index | – | 0.0%#152 | – | Artificial Analysis | |
| Terminal-Bench 2.1 (AA)not in index | 47.2%#124 | 47.6%#123 | – | Artificial Analysis |
Agents & tools
| Benchmark | no reasoning | thinking | Source | Trend |
|---|---|---|---|---|
| Terminal-Bench Hard | 29.6%#116 | 31.1%#106 | Artificial Analysis | |
| τ²-Bench Telecom (AA)not in index | 84.5%#102 | 93.6%#41 | Artificial Analysis | |
| GDPval-AA v2.1not in index | 10.4%#112 | 15.7%#103 | Artificial Analysis | |
| τ³-Banking (AA) | 10.3%#142 | 15.3%#117 | Artificial Analysis | |
| AutomationBenchnot in index | – | 5.5%#163 | Artificial Analysis | |
| GDP.pdf (AA)not in index | – | 3.4%#179 | Artificial Analysis | |
| AA-Briefcase v1.1not in index | – | 497#58 | Artificial Analysis |
Maths
| Benchmark | not stated | Source | Trend |
|---|---|---|---|
| Arena Maths | 1425#111 | Arena (LMArena) |
Knowledge
| Benchmark | no reasoning | thinking | Source | Trend |
|---|---|---|---|---|
| AA-Omniscience | -55.0#434 | -41.5#326 | Artificial Analysis | |
| AA-Omniscience: accuracynot in index | 19.1%#317 | 24.4%#244 | Artificial Analysis | |
| AA-Omniscience: non-hallucinationnot in index | 8.4%#462 | 12.9%#367 | Artificial Analysis | |
| Global-MMLU-Lite (AA)not in index | – | 89.1%#17 | Artificial Analysis |
Instruction following
| Benchmark | no reasoning | thinking | not stated | Source | Trend |
|---|---|---|---|---|---|
| IFBench | 50.8%#170 | 75.7%#27 | – | Artificial Analysis | |
| Arena Instruction Followingnot in indexeffort not stated | – | – | 1406#129 | Arena (LMArena) |
Human preference
| Benchmark | not stated | Source | Trend |
|---|---|---|---|
| Arena Text | 1417#128 | Arena (LMArena) | |
| Arena Creative Writingnot in index | 1366#151 | Arena (LMArena) | |
| Arena Multi-turnnot in index | 1420#130 | Arena (LMArena) | |
| Arena Longer Queriesnot in index | 1419#134 | Arena (LMArena) |
Multimodal
| Benchmark | no reasoning | thinking | not stated | Source | Trend |
|---|---|---|---|---|---|
| Arena Visioneffort not stated | – | – | 1245#69 | Arena (LMArena) | |
| MMMU-Pro | 70.3%#146 | 75.0%#106 | – | Artificial Analysis |
Long context
| Benchmark | no reasoning | thinking | Source | Trend |
|---|---|---|---|---|
| AA-LCR | 61.3%#273 | 76.3%#145 | Artificial Analysis | |
| MLCRnot in index | – | 7.8%#129 | Artificial Analysis |
Composite
| Benchmark | no reasoning | thinking | Source | Trend |
|---|---|---|---|---|
| AA Intelligence Index v4.3.2 | 17.8#260 | 15.6#279 | 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. | 103 tok/s | 0.95 s | $0.260 | $2.08 | 262k | – |
| NovitaAI | 93 tok/s | 1.18 s | $0.400 | $3.20 | 262k | bf16 |
| AtlasCloud | 86 tok/s | 1.12 s | $0.300 | $2.40 | 262k | fp8 |
| SiliconFlow | 45 tok/s | 1.40 s | $0.260 | $2.08 | 262k | fp8 |
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 | – | 6.6 s |
| Summarize a 30-page report | 12,000 / 600 | $0.0067 | – | 11.8 s |
| Code edit | 6,000 / 1,500 | $0.0072 | – | 27.6 s |
| Agentic coding session | 60,000 / 4,000 | $0.037 | – | 1.2 min |
| Structured extraction | 2,000 / 200 | $0.0014 | – | 4.8 s |
More from Alibaba
See also
Questions people ask
- What are Qwen3.5-122B-A10B's benchmark scores?
- As of 25 Oct 2026, Qwen3.5-122B-A10B scores 54.3 ±5.9 on the BenchLeader Index, strongest in maths (60) and weakest in coding (46). That rests on 6 independent benchmark results, and every score on this page links to its publisher.
- How does Qwen3.5-122B-A10B rank?
- Qwen3.5-122B-A10B is #256 of 785 ranked configurations on 25 Oct 2026. 333 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 Qwen3.5-122B-A10B cost per 1M tokens?
- As of 25 Oct 2026, Qwen3.5-122B-A10B costs $0.400 per 1M input tokens and $3.20 per 1M output tokens, $1.10 blended (three input tokens per output token).
- When was Qwen3.5-122B-A10B released?
- Alibaba released Qwen3.5-122B-A10B on 24 Feb 2026. BenchLeader has results for it at 3 reasoning-effort settings.
Data as of 25 Oct 2026. Compare these configurations.
Cite as: BenchLeader, “Qwen3.5-122B-A10B: benchmarks, pricing, speed and rank”, data as of 25 Oct 2026, https://www.benchleader.com/models/qwen3-5-122b-a10b.