# Deploy Qwen models on dedicated GPUs

56 Qwen models ready to deploy on GPU capacity reserved for you, from 596M to 2.45T parameters, covering chat & reasoning, coding, vision & multimodal, embeddings, rerankers, speech recognition, text-to-speech and safety & moderation. Private OpenAI-compatible endpoint, European GPUs, billed per GPU-hour.

| Model | Type | Params | Context | License | Page |
| --- | --- | --- | --- | --- | --- |
| Qwen3.8-27B | Chat & reasoning | 28B | 1M | Apache 2.0 | https://sference.com/models/Qwen/Qwen3.8-27B.md |
| Qwen3.8-Flash-Next | Chat & reasoning | 180B | 1M | Custom license | https://sference.com/models/Qwen/Qwen3.8-Flash-Next.md |
| Qwen3.6-35B-A3B | Chat & reasoning | 36B | 256K | Apache 2.0 | https://sference.com/models/Qwen/Qwen3.6-35B-A3B.md |
| Qwen3.5-9B | Chat & reasoning | 9.7B | 256K | Apache 2.0 | https://sference.com/models/Qwen/Qwen3.5-9B.md |
| Qwen3-TTS-12Hz-1.7B-CustomVoice | Text-to-speech | 1.9B | — | Apache 2.0 | https://sference.com/models/Qwen/Qwen3-TTS-12Hz-1.7B-CustomVoice.md |
| Qwen3-8B | Chat & reasoning | 8.2B | 40K | Apache 2.0 | https://sference.com/models/Qwen/Qwen3-8B.md |
| Qwen3-0.6B | Chat & reasoning | 752M | 40K | Apache 2.0 | https://sference.com/models/Qwen/Qwen3-0.6B.md |
| Qwen3.5-4B | Chat & reasoning | 4.7B | 256K | Apache 2.0 | https://sference.com/models/Qwen/Qwen3.5-4B.md |
| Qwen3-Embedding-0.6B | Embeddings | 596M | 32K | Apache 2.0 | https://sference.com/models/Qwen/Qwen3-Embedding-0.6B.md |
| Qwen3.6-27B | Chat & reasoning | 28B | 256K | Apache 2.0 | https://sference.com/models/Qwen/Qwen3.6-27B.md |
| Qwen3-ASR-1.7B | Speech recognition | 2.3B | — | Apache 2.0 | https://sference.com/models/Qwen/Qwen3-ASR-1.7B.md |
| Qwen3-VL-8B-Instruct | Vision & multimodal | 8.8B | 256K | Apache 2.0 | https://sference.com/models/Qwen/Qwen3-VL-8B-Instruct.md |
| Qwen3.5-35B-A3B | Chat & reasoning | 36B | 256K | Apache 2.0 | https://sference.com/models/Qwen/Qwen3.5-35B-A3B.md |
| Qwen3-4B-Instruct-2507 | Chat & reasoning | 4.0B | 256K | Apache 2.0 | https://sference.com/models/Qwen/Qwen3-4B-Instruct-2507.md |
| Qwen3-Coder-Next | Coding | 80B | 256K | Apache 2.0 | https://sference.com/models/Qwen/Qwen3-Coder-Next.md |
| Qwen3-Coder-30B-A3B-Instruct | Coding | 31B | 256K | Apache 2.0 | https://sference.com/models/Qwen/Qwen3-Coder-30B-A3B-Instruct.md |
| Qwen2.5-Omni-7B | Vision & multimodal | 11B | — | Apache 2.0 | https://sference.com/models/Qwen/Qwen2.5-Omni-7B.md |
| Qwen3.5-27B | Chat & reasoning | 28B | 256K | Apache 2.0 | https://sference.com/models/Qwen/Qwen3.5-27B.md |
| Qwen3.8-2.4T-A95B | Chat & reasoning | 2.45T | 1M | Custom license | https://sference.com/models/Qwen/Qwen3.8-2.4T-A95B.md |
| Qwen3-14B | Chat & reasoning | 15B | 40K | Apache 2.0 | https://sference.com/models/Qwen/Qwen3-14B.md |
| Qwen3-Embedding-8B | Embeddings | 7.6B | 40K | Apache 2.0 | https://sference.com/models/Qwen/Qwen3-Embedding-8B.md |
| Qwen3.5-0.8B | Chat & reasoning | 873M | 256K | Apache 2.0 | https://sference.com/models/Qwen/Qwen3.5-0.8B.md |
| Qwen3-1.7B | Chat & reasoning | 2.0B | 40K | Apache 2.0 | https://sference.com/models/Qwen/Qwen3-1.7B.md |
| Qwen3-30B-A3B | Chat & reasoning | 31B | 40K | Apache 2.0 | https://sference.com/models/Qwen/Qwen3-30B-A3B.md |
| Qwen3-235B-A22B | Chat & reasoning | 235B | 40K | Apache 2.0 | https://sference.com/models/Qwen/Qwen3-235B-A22B.md |
| Qwen3-Omni-30B-A3B-Instruct | Vision & multimodal | 35B | — | Apache 2.0 | https://sference.com/models/Qwen/Qwen3-Omni-30B-A3B-Instruct.md |
| Qwen3.5-122B-A10B | Chat & reasoning | 125B | 256K | Apache 2.0 | https://sference.com/models/Qwen/Qwen3.5-122B-A10B.md |
| Qwen3.5-2B | Chat & reasoning | 2.3B | 256K | Apache 2.0 | https://sference.com/models/Qwen/Qwen3.5-2B.md |
| Qwen3-VL-Embedding-8B | Embeddings | 8.1B | 256K | Apache 2.0 | https://sference.com/models/Qwen/Qwen3-VL-Embedding-8B.md |
| Qwen3-Reranker-0.6B | Rerankers | 596M | 40K | Apache 2.0 | https://sference.com/models/Qwen/Qwen3-Reranker-0.6B.md |
| Qwen3-Next-80B-A3B-Instruct | Chat & reasoning | 81B | 256K | Apache 2.0 | https://sference.com/models/Qwen/Qwen3-Next-80B-A3B-Instruct.md |
| Qwen3-VL-30B-A3B-Instruct | Vision & multimodal | 31B | 256K | Apache 2.0 | https://sference.com/models/Qwen/Qwen3-VL-30B-A3B-Instruct.md |
| Qwen3-30B-A3B-Instruct-2507 | Chat & reasoning | 31B | 256K | Apache 2.0 | https://sference.com/models/Qwen/Qwen3-30B-A3B-Instruct-2507.md |
| Qwen3.5-397B-A17B | Chat & reasoning | 403B | 256K | Apache 2.0 | https://sference.com/models/Qwen/Qwen3.5-397B-A17B.md |
| Qwen3-ASR-0.6B | Speech recognition | 938M | — | Apache 2.0 | https://sference.com/models/Qwen/Qwen3-ASR-0.6B.md |
| QwQ-32B | Chat & reasoning | 33B | 40K | Apache 2.0 | https://sference.com/models/Qwen/QwQ-32B.md |
| Qwen3-VL-Embedding-2B | Embeddings | 2.1B | 256K | Apache 2.0 | https://sference.com/models/Qwen/Qwen3-VL-Embedding-2B.md |
| Qwen3-32B | Chat & reasoning | 33B | 128K | Apache 2.0 | https://sference.com/models/Qwen/Qwen3-32B.md |
| Qwen3-Coder-480B-A35B-Instruct | Coding | 480B | 256K | Apache 2.0 | https://sference.com/models/Qwen/Qwen3-Coder-480B-A35B-Instruct.md |
| Qwen3-VL-4B-Instruct | Vision & multimodal | 4.4B | 256K | Apache 2.0 | https://sference.com/models/Qwen/Qwen3-VL-4B-Instruct.md |
| Qwen2.5-VL-72B-Instruct | Vision & multimodal | 73B | 125K | Custom license | https://sference.com/models/Qwen/Qwen2.5-VL-72B-Instruct.md |
| Qwen3-4B-Thinking-2507 | Chat & reasoning | 4.0B | 256K | Apache 2.0 | https://sference.com/models/Qwen/Qwen3-4B-Thinking-2507.md |
| Qwen3-VL-2B-Instruct | Vision & multimodal | 2.1B | 256K | Apache 2.0 | https://sference.com/models/Qwen/Qwen3-VL-2B-Instruct.md |
| Qwen3-235B-A22B-Instruct-2507 | Chat & reasoning | 235B | 256K | Apache 2.0 | https://sference.com/models/Qwen/Qwen3-235B-A22B-Instruct-2507.md |
| Qwen3-Reranker-4B | Rerankers | 4.0B | 40K | Apache 2.0 | https://sference.com/models/Qwen/Qwen3-Reranker-4B.md |
| Qwen2.5-VL-32B-Instruct | Vision & multimodal | 33B | 125K | Apache 2.0 | https://sference.com/models/Qwen/Qwen2.5-VL-32B-Instruct.md |
| Qwen3-VL-235B-A22B-Instruct | Vision & multimodal | 236B | 256K | Apache 2.0 | https://sference.com/models/Qwen/Qwen3-VL-235B-A22B-Instruct.md |
| Qwen3-30B-A3B-Thinking-2507 | Chat & reasoning | 31B | 256K | Apache 2.0 | https://sference.com/models/Qwen/Qwen3-30B-A3B-Thinking-2507.md |
| Qwen3-VL-Reranker-2B | Rerankers | 2.1B | 256K | Apache 2.0 | https://sference.com/models/Qwen/Qwen3-VL-Reranker-2B.md |
| Qwen3-Embedding-4B | Embeddings | 4.0B | 40K | Apache 2.0 | https://sference.com/models/Qwen/Qwen3-Embedding-4B.md |
| Qwen3-Next-80B-A3B-Thinking | Chat & reasoning | 81B | 256K | Apache 2.0 | https://sference.com/models/Qwen/Qwen3-Next-80B-A3B-Thinking.md |
| Qwen3Guard-Gen-8B | Safety & moderation | 8.2B | 32K | Apache 2.0 | https://sference.com/models/Qwen/Qwen3Guard-Gen-8B.md |
| Qwen3-Reranker-8B | Rerankers | 8.2B | 40K | Apache 2.0 | https://sference.com/models/Qwen/Qwen3-Reranker-8B.md |
| Qwen3-VL-235B-A22B-Thinking | Vision & multimodal | 236B | 256K | Apache 2.0 | https://sference.com/models/Qwen/Qwen3-VL-235B-A22B-Thinking.md |
| Qwen3-235B-A22B-Thinking-2507 | Chat & reasoning | 235B | 125K | Apache 2.0 | https://sference.com/models/Qwen/Qwen3-235B-A22B-Thinking-2507.md |
| Qwen3-VL-Reranker-8B | Rerankers | 8.8B | 256K | Apache 2.0 | https://sference.com/models/Qwen/Qwen3-VL-Reranker-8B.md |

## FAQ

### Which Qwen models can I deploy on sference?

56 today, most requested first: Qwen3.8-27B, Qwen3.8-Flash-Next, Qwen3.6-35B-A3B, Qwen3.5-9B, Qwen3-TTS-12Hz-1.7B-CustomVoice, Qwen3-8B, Qwen3-0.6B and Qwen3.5-4B, and 48 more. Any other Qwen checkpoint or a fine-tune of one can be requested with the form on this page.

### How many GPUs do Qwen models need?

It depends on the model: the lineup runs from 596M to 2.45T parameters. Each model page lists the checkpoint size and the minimum footprint on 80 GB and 141 GB GPUs. You don't pick hardware on sference: you choose a performance tier and we size the deployment.

### Can I use Qwen models commercially?

53 of the 56 listed models are under licenses that allow commercial use (Apache 2.0 and Custom license across the lineup). Each model page summarizes its license and links the full text.

### How is a dedicated deployment priced?

Per GPU-hour on a monthly commitment, sized to your throughput and latency targets.

Request a deployment: https://sference.com/models/Qwen#request. All makers: https://sference.com/dedicated.md