Alibaba Cloud / Qwen Team model product
Qwen3.6 35B A3B
6 series, a multimodal Mixture-of-Experts model with 35B total parameters and 3B activated.
Updated Aug 12, 2026. Default version: Qwen3.6-35B-A3B
LLMBoard score55.3Qwen3.6-35B-A3B
Coverage80%49 benchmark families
Context window262.1KTokens
Official input price$0.248Alibaba API
Qwen3.6 35B A3B Specifications
Technical details for the model's default version.
- Version
- Qwen3.6-35B-A3B
- Released
- Apr 16, 2026
- Knowledge cutoff
- Unknown
- Parameters
- 35B
- Context window
- 262.1K
- Max output
- 65.5K
- Inputs
- audio, image, text, video
- Outputs
- text
- Open weights
- Yes
- License
- Apache 2.0
Qwen3.6 35B A3B Capability Profile
This profile uses the model's current scored version. Arena ratings and prices are shown separately.
Qwen3.6 35B A3B Benchmark Results
Benchmark scores for Qwen3.6-35B-A3B.
| Hallusion Bench | 69.8% | 02 | 16 | 93.3% | C | |
| MMBench-V1.1 | 92.8% | 02 | 18 | 94.1% | C | |
| CC-OCR | 81.9% | 03 | 18 | 88.2% | C | |
| ODinW | 50.8% | 03 | 16 | 86.7% | C | |
| ZEROBench-Sub | 0.344 points | 03 | 5 | 50.0% | C | |
| DeepPlanning | 25.9% | 04 | 9 | 62.5% | C | |
| EmbSpatialBench | 0.843 points | 04 | 8 | 57.1% | C | |
| OmniDocBench 1.5 | 89.9% | 04 | 17 | 81.3% | C | |
| RefCOCO-avg | 0.92 points | 04 | 7 | 50.0% | C | |
| TAU3-Bench | 67.2% | 04 | 5 | 25.0% | C | |
| VideoMME w/o sub. | 82.5% | 04 | 10 | 66.7% | C | |
| ZClawBench | 52.6% | 04 | 4 | 0.0% | C | |
| MathVista-Mini | 86.4% | 05 | 23 | 81.8% | C | |
| MLVU | 86.2% | 05 | 10 | 55.6% | C | |
| MVBench | 74.6% | 05 | 17 | 75.0% | C | |
| RefSpatialBench | 0.643 points | 05 | 6 | 20.0% | C | |
| RealWorldQA | 85.3% | 06 | 26 | 80.0% | C | |
| VITA-Bench | 35.6% | 06 | 10 | 44.4% | C | |
| VideoMME w sub. | 86.6% | 07 | 10 | 33.3% | C | |
| AI2D | 92.7% | 08 | 32 | 77.4% | C | |
| MCP-Mark | 37.0% | 08 | 8 | 0.0% | C | |
| SkillsBench | 28.7% | 08 | 8 | 0.0% | C | |
| WideSearch | 60.1% | 08 | 9 | 12.5% | C | |
| AIME 2026 | 92.7% | 09 | 18 | 52.9% | C | |
| C-Eval | 90.0% | 09 | 18 | 52.9% | C | |
| HMMT25 | 89.1% | 09 | 25 | 66.7% | C | |
| LVBench | 71.4% | 09 | 24 | 65.2% | C | |
| HMMT Feb 26 | 83.6% | 10 | 11 | 10.0% | C | |
| SimpleVQA | 0.589 points | 10 | 13 | 25.0% | C | |
| MMMU | 81.7% | 11 | 63 | 83.9% | C | |
| VideoMMMU | 83.7% | 11 | 26 | 60.0% | C | |
| MMLU-Redux | 93.3% | 12 | 48 | 76.6% | C | |
| SuperGPQA | 64.7% | 12 | 34 | 66.7% | C | |
| Claw-Eval | 50.0% | 13 | 13 | 0.0% | C | |
| NL2Repo | 29.4% | 14 | 14 | 0.0% | C | |
| HMMT 2025 | 90.7% | 16 | 33 | 53.1% | C | |
| IMO-AnswerBench | 78.9% | 16 | 19 | 16.7% | C | |
| MMLU-Pro | 85.2% | 16 | 129 | 88.3% | C | |
| LiveCodeBench v6 | 80.4% | 20 | 53 | 63.5% | C | |
| SWE-bench Multilingual | 67.2% | 22 | 34 | 36.4% | C | |
| MCP Atlas | 62.8% | 23 | 31 | 26.7% | C | |
| CharXiv-R | 78.0% | 25 | 48 | 48.9% | C | |
| Toolathlon | 26.9% | 31 | 31 | 0.0% | C | |
| MMMU-Pro | 75.3% | 32 | 66 | 52.3% | C | |
| Terminal-Bench 2.0 | 51.5% | 33 | 49 | 33.3% | C | |
| SWE-Bench Pro | 49.5% | 44 | 45 | 2.3% | C | |
| SWE-Bench Verified | 73.4% | 45 | 105 | 57.7% | C | |
| GPQA | 86.0% | 47 | 234 | 80.3% | C | |
| Humanity's Last Exam | 21.4% | 58 | 93 | 38.0% | C |
Qwen3.6 35B A3B Arena Results
Preference and agent-evaluation results for the default version.
No Arena results
The default version does not have a matching Arena result yet.
Qwen3.6 35B A3B Runtime Performance
Provider-specific output speed and catalog latency for Qwen3.6-35B-A3B. 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.
Qwen3.6 35B A3B Pricing
Official vendor API pricing appears first, followed by individual provider offers.
- Official API
- $0.248 input, $1.49 output per 1M
- Official provider
- Alibaba
- Lowest third-party
- From $0.07 input, $0.42 output per 1M via EmpirioLabs AI
- Tracked offerings
- 25
| QVAC | qwen3.6-35b-a3b | global | N/A | N/A | 262.1K | |
| Umans AI Coding Plan | umans-qwen3.6-35b-a3b | global | N/A | N/A | 262.1K | |
| EmpirioLabs AI | qwen3-6-35b-a3b | global | $0.07 | $0.42 | 131.1K | |
| SaladCloud AI Gateway | qwen3.6-35b-a3b | global | $0.09 | $0.60 | 262.1K | |
| Venice AI | qwen3-6-35b-a3b | global | $0.10 | $1 | 256K | |
| Deep Infra | Qwen/Qwen3.6-35B-A3B | global | $0.10 | $0.95 | 262.1K | |
| NanoGPT | qwen/Qwen3.6-35B-A3B | global | $0.112 | $0.80 | 262.1K | |
| Pioneer | Qwen/Qwen3.6-35B-A3B | global | $0.14 | $1 | 262.1K | |
| AKI.IO | qwen3.6-35b | global | $0.15 | $0.50 | 256K | |
| Kilo Gateway | qwen/qwen3.6-35b-a3b | global | $0.15 | $1 | 262.1K | |
| OpenRouter | qwen/qwen3.6-35b-a3b | global | $0.15 | $1 | 262.1K | |
| Hugging Face | Qwen/Qwen3.6-35B-A3B | global | $0.15 | $0.95 | 262.1K | |
| Cortecs | qwen3.6-35b-a3b | global | $0.167 | $0.557 | 262K | |
| Zenifra | alibaba/qwen3.6-35b-a3b | global | $0.19 | $0.48 | 262.1K | |
| SiliconFlow | Qwen/Qwen3.6-35B-A3B | global | $0.20 | $1.6 | 262.1K | |
| LLM Gateway | qwen3.6-35b-a3b | global | $0.248 | $1.49 | 262.1K | |
| OrcaRouter | qwen/qwen3.6-35b-a3b | global | $0.248 | $1.49 | 262.1K | |
| Merge Gateway | qwen/qwen3.6-35b-a3b | global | $0.248 | $1.49 | 262.1K | |
| Alibaba | qwen3.6-35b-a3b | global | $0.248 | $1.49 | 262.1K | |
| Weights & Biases | Qwen/Qwen3.6-35B-A3B | global | $0.25 | $1.25 | 262.1K | |
| Scaleway | qwen3.6-35b-a3b | global | $0.25 | $1.5 | 128K | |
| Neuralwatt | Qwen/Qwen3.6-35B-A3B | global | $0.29 | $1.15 | 131.1K | |
| GreenPT | qwen3.6-35b-a3b | global | $0.342 | $2.05 | 262.1K | |
| evroc | Qwen/Qwen3.6-35B-A3B-FP8 | global | $0.345 | $1.38 | 262.1K | |
| LLMTR | qwen3-6-35b | global | $5 | $10 | 16.4K |
Official prices use only the vendor's configured official Provider and positive standard USD PAYG rates. Third-party offers remain explicitly labeled.
Qwen3.6 35B A3B Versions
Available versions of this model. The score column identifies the version used in the overall ranking.
| Qwen3.6-35B-A3B | 55.3 | 35B | 262.1K | 65.5K | Yes | Apache 2.0 |
What is Qwen3.6 35B A3B?
Key information about Qwen3.6 35B A3B and its available data.
6 series, a multimodal Mixture-of-Experts model with 35B total parameters and 3B activated.
It pairs a vision encoder with a hybrid 40-layer language model that interleaves Gated DeltaNet linear-attention blocks and Gated Attention blocks (10 × (3 × DeltaNet + 1 × Attention)) over 256 experts (8 routed + 1 shared, expert dim 512).
The release prioritizes stability and real-world utility, with substantial gains in agentic coding (frontend workflows, repo-level reasoning) and a new option to preserve reasoning context across turns. Native context length is 262K tokens, extensible to ~1M via YaRN, and the model thinks by default.
Data as of 2026-08-11.
Qwen3.6 35B A3B vs nearby models
Open a comparison with the three ranked models immediately above and below this model.
Models similar to Qwen3.6 35B A3B
Recommendations prioritize the same model type and family, then the closest LLMBoard score.
AC
Qwen3.5 27B
Alibaba Cloud / Qwen Team
AC
Qwen3.5 35B A3B
Alibaba Cloud / Qwen Team
AC
Qwen3.5 122B A10B
Alibaba Cloud / Qwen Team
AC
Qwen3.6 27B
Alibaba Cloud / Qwen Team
AC
Qwen3 235B A22B Thinking
Alibaba Cloud / Qwen Team
AC
Qwen3 VL 235B A22B Thinking
Alibaba Cloud / Qwen Team
FAQ
Common questions about Qwen3.6 35B A3B.
When was Qwen3.6 35B A3B released?
Qwen3.6 35B A3B's default version was released on Apr 16, 2026.
How much does Qwen3.6 35B A3B cost?
Qwen3.6 35B A3B's official API price is $0.248 per million input tokens and $1.49 per million output tokens via Alibaba. The lowest tracked third-party offer starts at $0.07 input and $0.42 output via EmpirioLabs AI.
Who created Qwen3.6 35B A3B?
Qwen3.6 35B A3B was created by Alibaba Cloud / Qwen Team.
What is the context window for Qwen3.6 35B A3B?
The default version has a 262.1K token context window.
Is Qwen3.6 35B A3B open weight?
Yes. The default version is marked as open weight under Apache 2.0.
How many API providers offer Qwen3.6 35B A3B?
25 provider offerings are linked to the default version.
What models should I compare Qwen3.6 35B A3B with?
Nearby ranked alternatives include Grok 4, Claude Sonnet 4.5, MiniMax M2.5.