Alibaba Cloud / Qwen Team model product
Qwen3.5 122B A10B
5-122B-A10B is a multimodal Mixture-of-Experts model with 122 billion total parameters and 10 billion activated parameters.
Updated Aug 12, 2026. Default version: Qwen3.5-122B-A10B
LLMBoard score59.2Qwen3.5-122B-A10B
Coverage100%81 benchmark families
Context window262.1KTokens
Official input price$0.40Alibaba API
Qwen3.5 122B A10B Specifications
Technical details for the model's default version.
- Version
- Qwen3.5-122B-A10B
- Released
- Feb 24, 2026
- Knowledge cutoff
- Unknown
- Parameters
- 122B
- Context window
- 262.1K
- Max output
- 65.5K
- Inputs
- audio, image, text, video
- Outputs
- text
- Open weights
- Yes
- License
- Apache 2.0
Qwen3.5 122B A10B Capability Profile
This profile uses the model's current scored version. Arena ratings and prices are shown separately.
Qwen3.5 122B A10B Benchmark Results
Benchmark scores for Qwen3.5-122B-A10B.
| FullStackBench en | 62.6% | 01 | 3 | 100.0% | C | |
| FullStackBench zh | 58.7% | 01 | 3 | 100.0% | C | |
| MMBench-V1.1 | 92.8% | 01 | 18 | 100.0% | C | |
| MVBench | 76.6% | 01 | 17 | 100.0% | C | |
| Nuscene | 15.4% | 01 | 3 | 100.0% | C | |
| PMC-VQA | 63.3% | 01 | 3 | 100.0% | C | |
| SlakeVQA | 81.6% | 01 | 4 | 100.0% | C | |
| SUNRGBD | 0.362 points | 01 | 4 | 100.0% | C | |
| VideoMME w/o sub. | 83.9% | 01 | 10 | 100.0% | C | |
| ZEROBench-Sub | 0.362 points | 01 | 5 | 100.0% | C | |
| AndroidWorld_SR | 66.4% | 02 | 8 | 85.7% | C | |
| BrowseComp-zh | 69.9% | 02 | 13 | 91.7% | C | |
| MedXpertQA | 67.3% | 02 | 12 | 90.9% | C | |
| MLVU | 87.3% | 02 | 10 | 88.9% | C | |
| MMStar | 82.9% | 02 | 22 | 95.2% | C | |
| MMVU | 74.7% | 02 | 4 | 66.7% | C | |
| OCRBench | 92.1% | 02 | 22 | 95.2% | C | |
| DynaMath | 85.9% | 03 | 7 | 66.7% | C | |
| Hypersim | 0.127 points | 03 | 4 | 33.3% | C | |
| LingoQA | 80.8% | 03 | 4 | 33.3% | C | |
| MathVista-Mini | 87.4% | 03 | 23 | 90.9% | C | |
| RefSpatialBench | 0.693 points | 03 | 6 | 60.0% | C | |
| VideoMME w sub. | 87.3% | 03 | 10 | 77.8% | C | |
| BFCL-V4 | 72.2% | 04 | 13 | 75.0% | C | |
| CC-OCR | 81.8% | 04 | 18 | 82.3% | C | |
| C-Eval | 91.9% | 04 | 18 | 82.3% | C | |
| CodeForces | 85.1% | 04 | 16 | 80.0% | C | |
| Hallusion Bench | 67.6% | 04 | 16 | 80.0% | C | |
| MMLongBench-Doc | 0.59 points | 04 | 5 | 25.0% | C | |
| NOVA-63 | 58.6% | 04 | 11 | 70.0% | C | |
| OJBench | 39.5% | 04 | 9 | 62.5% | C | |
| TIR-Bench | 53.2% | 04 | 4 | 0.0% | C | |
| VLMsAreBlind | 96.7% | 04 | 4 | 0.0% | C | |
| AI2D | 93.3% | 05 | 32 | 87.1% | C | |
| CountBench | 0.97 points | 05 | 6 | 20.0% | C | |
| DeepPlanning | 24.1% | 05 | 9 | 50.0% | C | |
| EmbSpatialBench | 0.839 points | 05 | 8 | 42.9% | C | |
| HMMT25 | 90.3% | 05 | 25 | 83.3% | C | |
| OmniDocBench 1.5 | 89.8% | 05 | 17 | 75.0% | C | |
| RefCOCO-avg | 0.913 points | 05 | 7 | 33.3% | C | |
| Seal-0 | 44.1% | 05 | 6 | 20.0% | C | |
| V* | 93.2% | 05 | 7 | 33.3% | C | |
| IFEval | 93.4% | 06 | 65 | 92.2% | C | |
| Include | 82.8% | 06 | 31 | 83.3% | C | |
| LVBench | 74.4% | 06 | 24 | 78.3% | C | |
| MAXIFE | 87.9% | 06 | 11 | 50.0% | C | |
| MMLU-ProX | 82.2% | 06 | 32 | 83.9% | C | |
| MMLU-Redux | 94.0% | 06 | 48 | 89.4% | C | |
| MMMU | 83.9% | 06 | 63 | 91.9% | C | |
| PolyMATH | 68.9% | 06 | 23 | 77.3% | C | |
| AA-LCR | 66.9% | 07 | 16 | 60.0% | C | |
| Global PIQA | 88.4% | 07 | 13 | 50.0% | C | |
| MathVision | 86.2% | 07 | 32 | 80.7% | C | |
| RealWorldQA | 85.1% | 07 | 26 | 76.0% | C | |
| SuperGPQA | 67.1% | 07 | 34 | 81.8% | C | |
| VITA-Bench | 33.6% | 07 | 10 | 33.3% | C | |
| WideSearch | 60.5% | 07 | 9 | 25.0% | C | |
| ZEROBench | 0.09 points | 07 | 9 | 25.0% | C | |
| BabyVision | 40.2% | 08 | 9 | 12.5% | C | |
| Multi-Challenge | 61.5% | 08 | 29 | 75.0% | C | |
| ScreenSpot Pro | 70.4% | 08 | 25 | 70.8% | C | |
| SimpleVQA | 0.617 points | 08 | 13 | 41.7% | C | |
| ODinW | 44.5% | 09 | 16 | 46.7% | C | |
| WMT24++ | 78.3% | 09 | 23 | 63.6% | C | |
| IFBench | 76.1% | 10 | 29 | 67.9% | C | |
| MMLU-Pro | 86.7% | 10 | 129 | 93.0% | C | |
| ERQA | 62.0% | 11 | 23 | 54.5% | C | |
| LongBench v2 | 60.2% | 11 | 17 | 37.5% | C | |
| t2-bench | 79.5% | 13 | 23 | 45.5% | C | |
| HMMT 2025 | 91.4% | 15 | 33 | 56.3% | C | |
| VideoMMMU | 82.0% | 15 | 26 | 44.0% | C | |
| OSWorld-Verified | 58.0% | 20 | 23 | 13.6% | C | |
| LiveCodeBench v6 | 78.9% | 22 | 53 | 59.6% | C | |
| MMMLU | 86.7% | 22 | 49 | 56.3% | C | |
| Humanity's Last Exam | 47.5% | 24 | 93 | 75.0% | C | |
| MMMU-Pro | 76.9% | 25 | 66 | 63.1% | C | |
| CharXiv-R | 77.2% | 29 | 48 | 40.4% | C | |
| BrowseComp | 63.8% | 32 | 58 | 45.6% | C | |
| Terminal-Bench 2.0 | 49.4% | 36 | 49 | 27.1% | C | |
| GPQA | 86.6% | 44 | 234 | 81.5% | C | |
| SWE-Bench Verified | 72.0% | 53 | 105 | 50.0% | C |
Qwen3.5 122B A10B Arena Results
Preference and agent-evaluation results for the default version.
| vision | overall | 53 | 1245.6 | 12,437 | N/A | |
| vision style control | overall | 55 | 1226.9 | 12,437 | N/A | |
| webdev | overall | 77 | 1357.6 | 7,845 | N/A | |
| text factuality | overall | 99 | 1428.5 | 28,267 | N/A | |
| text | overall | 103 | 1418.0 | 28,459 | N/A | |
| text style control | overall | 112 | 1417.3 | 28,459 | N/A |
Qwen3.5 122B A10B Runtime Performance
Provider-specific output speed and catalog latency for Qwen3.5-122B-A10B. Runtime does not affect the capability score.
No runtime data
Browse runtime rankingsNo provider-specific speed or latency record is linked to the default version yet.
Output Speed is generated output tokens received per second. Catalog latency is reported separately from observed provider TTFT.
Qwen3.5 122B A10B Pricing
Official vendor API pricing appears first, followed by individual provider offers.
- Official API
- $0.40 input, $3.2 output per 1M
- Official provider
- Alibaba
- Lowest third-party
- From $0.115 input, $0.917 output per 1M via EmpirioLabs AI
- Tracked offerings
- 18
| Nvidia | qwen/qwen3.5-122b-a10b | global | N/A | N/A | 262.1K | |
| EmpirioLabs AI | qwen3-5-122b-a10b | global | $0.115 | $0.917 | 256K | |
| OrcaRouter | qwen/qwen3.5-122b-a10b | global | $0.115 | $0.917 | 262.1K | |
| Merge Gateway | qwen/qwen3.5-122b-a10b | global | $0.115 | $0.917 | 131.1K | |
| Neon | qwen35-122b-a10b | global | $0.22 | $2.2 | 262.1K | |
| SiliconFlow | Qwen/Qwen3.5-122B-A10B | global | $0.26 | $2.08 | 262.1K | |
| Kilo Gateway | qwen/qwen3.5-122b-a10b | global | $0.26 | $2.08 | 262.1K | |
| Ofox | bailian/qwen3.5-122b-a10b | global | $0.29 | $2.29 | 256K | |
| OpenRouter | qwen/qwen3.5-122b-a10b | global | $0.29 | $2.4 | 262.1K | |
| Deep Infra | Qwen/Qwen3.5-122B-A10B | global | $0.29 | $2.4 | 262.1K | |
| NEAR AI Cloud | Qwen/Qwen3.5-122B-A10B | global | $0.40 | $3.2 | 131.1K | |
| Mixlayer | qwen/qwen3.5-122b-a10b | global | $0.40 | $3.2 | 262.1K | |
| Alibaba | qwen3.5-122b-a10b | global | $0.40 | $3.2 | 262.1K | |
| NovitaAI | qwen/qwen3.5-122b-a10b | global | $0.40 | $3.2 | 262.1K | |
| Hugging Face | Qwen/Qwen3.5-122B-A10B | global | $0.40 | $3.2 | 262.1K | |
| NanoGPT | qwen3.5-122b-a10b | global | $0.437 | $3.5 | 131.1K | |
| Cortecs | qwen3.5-122b-a10b | global | $0.495 | $3.46 | 262.1K | |
| TensorX | qwen/qwen3.5-122b-a10b | global | $0.50 | $3.5 | 262.1K |
Official prices use only the vendor's configured official Provider and positive standard USD PAYG rates. Third-party offers remain explicitly labeled.
Qwen3.5 122B A10B Versions
Available versions of this model. The score column identifies the version used in the overall ranking.
| Qwen3.5-122B-A10B | 59.2 | 122B | 262.1K | 65.5K | Yes | Apache 2.0 |
What is Qwen3.5 122B A10B?
Key information about Qwen3.5 122B A10B and its available data.
5-122B-A10B is a multimodal Mixture-of-Experts model with 122 billion total parameters and 10 billion activated parameters. It combines strong reasoning, coding, long-context, and visual understanding performance with production-friendly efficiency and a native 262K context window.
Data as of 2026-08-11.
Qwen3.5 122B A10B vs nearby models
Open a comparison with the three ranked models immediately above and below this model.
Models similar to Qwen3.5 122B A10B
Recommendations prioritize the same model type and family, then the closest LLMBoard score.
AC
Qwen3.6 27B
Alibaba Cloud / Qwen Team
AC
Qwen3.5 27B
Alibaba Cloud / Qwen Team
AC
Qwen3.6 35B A3B
Alibaba Cloud / Qwen Team
AC
Qwen3.5 397B A17B
Alibaba Cloud / Qwen Team
AC
Qwen3.5 35B A3B
Alibaba Cloud / Qwen Team
AC
Qwen3.6 Plus
Alibaba Cloud / Qwen Team
FAQ
Common questions about Qwen3.5 122B A10B.
When was Qwen3.5 122B A10B released?
Qwen3.5 122B A10B's default version was released on Feb 24, 2026.
How much does Qwen3.5 122B A10B cost?
Qwen3.5 122B A10B's official API price is $0.40 per million input tokens and $3.2 per million output tokens via Alibaba. The lowest tracked third-party offer starts at $0.115 input and $0.917 output via EmpirioLabs AI.
Who created Qwen3.5 122B A10B?
Qwen3.5 122B A10B was created by Alibaba Cloud / Qwen Team.
What is the context window for Qwen3.5 122B A10B?
The default version has a 262.1K token context window.
Is Qwen3.5 122B A10B open weight?
Yes. The default version is marked as open weight under Apache 2.0.
How many API providers offer Qwen3.5 122B A10B?
18 provider offerings are linked to the default version.
What models should I compare Qwen3.5 122B A10B with?
Nearby ranked alternatives include Step 3.5 Flash, Gemma 4 31B, GLM 4.7.