Alibaba Cloud / Qwen Team model product
Qwen3.5 9B
5-9B is a 9 billion parameter vision-language model using Gated DeltaNet hybrid architecture with a 3:1 ratio of linear attention to full softmax attention.
Updated Aug 12, 2026. Default version: Qwen3.5-9B
LLMBoard score41.4Qwen3.5-9B
Coverage80%25 benchmark families
Context window262.1KTokens
Official input priceN/AOfficial price unavailable
Qwen3.5 9B Specifications
Technical details for the model's default version.
- Version
- Qwen3.5-9B
- Released
- Mar 2, 2026
- Knowledge cutoff
- Unknown
- Parameters
- 9B
- Context window
- 262.1K
- Max output
- 65.5K
- Inputs
- image, text, video
- Outputs
- text
- Open weights
- Yes
- License
- Apache 2.0
Qwen3.5 9B Capability Profile
This profile uses the model's current scored version. Arena ratings and prices are shown separately.
Qwen3.5 9B Benchmark Results
Benchmark scores for Qwen3.5-9B.
| BFCL-V4 | 0.7 | 7 | 13 | 50.0% | C | |
| DeepPlanning | 0.2 | 8 | 9 | 12.5% | C | |
| MAXIFE | 0.8 | 8 | 11 | 30.0% | C | |
| NOVA-63 | 0.6 | 8 | 11 | 30.0% | C | |
| PolyMATH | 0.6 | 9 | 23 | 63.6% | C | |
| VITA-Bench | 0.3 | 9 | 10 | 11.1% | C | |
| AA-LCR | 0.6 | 10 | 16 | 40.0% | C | |
| Global PIQA | 0.8 | 10 | 13 | 25.0% | C | |
| C-Eval | 0.9 | 11 | 18 | 41.2% | C | |
| HMMT25 | 0.8 | 12 | 25 | 54.2% | C | |
| IFEval | 0.9 | 12 | 65 | 82.8% | C | |
| WMT24++ | 0.7 | 12 | 23 | 50.0% | C | |
| LongBench v2 | 0.6 | 13 | 17 | 25.0% | C | |
| Multi-Challenge | 0.5 | 13 | 29 | 57.1% | C | |
| t2-bench | 0.8 | 14 | 23 | 40.9% | C | |
| Include | 0.8 | 16 | 31 | 50.0% | C | |
| MMLU-ProX | 0.8 | 17 | 32 | 48.4% | C | |
| SuperGPQA | 0.6 | 20 | 34 | 42.4% | C | |
| MMLU-Redux | 0.9 | 22 | 48 | 55.3% | C | |
| HMMT 2025 | 0.8 | 24 | 33 | 28.1% | C | |
| IFBench | 0.6 | 24 | 29 | 17.9% | C | |
| LiveCodeBench v6 | 0.7 | 36 | 53 | 32.7% | C | |
| MMLU-Pro | 0.8 | 37 | 129 | 71.9% | C | |
| MMMLU | 0.8 | 39 | 49 | 20.8% | C | |
| GPQA | 0.8 | 76 | 234 | 67.8% | C |
Qwen3.5 9B 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.
Qwen3.5 9B Pricing
Official vendor API pricing appears first, followed by individual provider offers.
- Official API
- N/A
- Official provider
- N/A
- Lowest third-party
- From $0.04 input, $0.15 output per 1M via CrofAI
- Tracked offerings
- 19
| QVAC | qwen3.5-9b | global | N/A | N/A | 32.8K | |
| CrofAI | qwen3.5-9b | global | $0.04 | $0.15 | 262.1K | |
| NanoGPT | qwen/qwen3.5-9b | global | $0.05 | $0.15 | 256K | |
| EmpirioLabs AI | qwen3-5-9b | global | $0.09 | $0.13 | 262.1K | |
| Merge Gateway | qwen/qwen3.5-9b | global | $0.09 | $0.13 | 262.1K | |
| LLM Gateway | qwen3.5-9b | global | $0.10 | $0.15 | 262.1K | |
| Venice AI | qwen3-5-9b | global | $0.10 | $0.15 | 256K | |
| Mixlayer | qwen/qwen3.5-9b | global | $0.10 | $0.40 | 262.1K | |
| Kilo Gateway | qwen/qwen3.5-9b | global | $0.10 | $0.15 | 262.1K | |
| OpenRouter | qwen/qwen3.5-9b | global | $0.10 | $0.15 | 262.1K | |
| Deep Infra | Qwen/Qwen3.5-9B | global | $0.10 | $0.15 | 262.1K | |
| Cortecs | qwen3.5-9b | global | $0.111 | $0.167 | 262.1K | |
| OVHcloud AI Endpoints | qwen3.5-9b | global | $0.12 | $0.18 | 262.1K | |
| Regolo AI | qwen3.5-9b | global | $0.15 | $0.60 | 262.1K | |
| TensorX | qwen/qwen3.5-9b | global | $0.15 | $0.20 | 262.1K | |
| routing.run | qwen3.5-9b | global | $0.16 | $0.48 | 262.1K | |
| Together AI | Qwen/Qwen3.5-9B | global | $0.17 | $0.25 | 262.1K | |
| Hugging Face | Qwen/Qwen3.5-9B | global | $0.17 | $0.25 | 262.1K | |
| Pioneer | Qwen/Qwen3.5-9B | global | $0.30 | $0.30 | 32.8K |
Official prices use only the vendor's configured official Provider and positive standard USD PAYG rates. Third-party offers remain explicitly labeled.
Qwen3.5 9B Versions
Available versions of this model. The score column identifies the version used in the overall ranking.
| Qwen3.5-9B | 41.4 | 9B | 262.1K | 65.5K | Yes | Apache 2.0 |
What is Qwen3.5 9B?
Key information about Qwen3.5 9B and its available data.
5-9B is a 9 billion parameter vision-language model using Gated DeltaNet hybrid architecture with a 3:1 ratio of linear attention to full softmax attention. It supports 262K native context length and delivers strong performance across knowledge, reasoning, coding, and multilingual tasks.
Data as of 2026-08-11.
Qwen3.5 9B vs nearby models
Open a comparison with the three ranked models immediately above and below this model.
Models similar to Qwen3.5 9B
Recommendations prioritize the same model type and family, then the closest LLMBoard score.
AC
Qwen3 Next 80B A3B Thinking
Alibaba Cloud / Qwen Team
AC
Qwen3 Max
Alibaba Cloud / Qwen Team
AC
Qwen3 VL 32B Thinking
Alibaba Cloud / Qwen Team
AC
Qwen3 235B A22B
Alibaba Cloud / Qwen Team
AC
Qwen3 VL 235B A22B
Alibaba Cloud / Qwen Team
AC
Qwen3 VL 235B A22B Thinking
Alibaba Cloud / Qwen Team
FAQ
Common questions about Qwen3.5 9B.
When was Qwen3.5 9B released?
Qwen3.5 9B's default version was released on Mar 2, 2026.
How much does Qwen3.5 9B cost?
No official standard PAYG price is currently available for Qwen3.5 9B. The lowest tracked third-party offer starts at $0.04 input and $0.15 output via CrofAI.
Who created Qwen3.5 9B?
Qwen3.5 9B was created by Alibaba Cloud / Qwen Team.
What is the context window for Qwen3.5 9B?
The default version has a 262.1K token context window.
Is Qwen3.5 9B open weight?
Yes. The default version is marked as open weight under Apache 2.0.
How many API providers offer Qwen3.5 9B?
19 provider offerings are linked to the default version.
What models should I compare Qwen3.5 9B with?
Nearby ranked alternatives include GPT-5.3, Command A+, GLM 4.5 Air.