Alibaba / Qwen
Latest releases
Alibaba / Qwen
- Qwen-Audio 3.0 Realtime Flash2026-07-14 · Model releaseOriginal release evidenceDate basis: Official Alibaba Cloud newly released models table dateChecked: 2026-07-15
- Qwen-Audio 3.0 Realtime Plus2026-07-14 · Model releaseOriginal release evidenceDate basis: Official Alibaba Cloud newly released models table dateChecked: 2026-07-15
- Qwen-Audio 3.0 TTS Flash2026-07-14 · Model releaseOriginal release evidenceDate basis: Official Alibaba Cloud newly released models table dateChecked: 2026-07-15
- Qwen-Audio 3.0 TTS Plus2026-07-14 · Model releaseOriginal release evidenceDate basis: Official Alibaba Cloud newly released models table dateChecked: 2026-07-15
Current models
Models, versions, and checkable fields
| Model | Version | Release | OpenRouter | Pricing | Benchmark | Public Signals |
|---|---|---|---|---|---|---|
| Qwen3.7-Max Alibaba Cloud's flagship Qwen3.7 API model for complex reasoning, coding, and million-token context workloads. | qwen3.7-max | 2026-05-21 | OpenRouter listed | OpenRouter · 2026-07-15: $1.475 / $4.425 input / output, per 1M tokens | No reliable public source | No reliable public source |
| Qwen3.7-Plus The cost-balanced Qwen3.7 general API model with text, image, and video understanding. | qwen3.7-plus | 2026-06-02 | OpenRouter listed | OpenRouter · 2026-07-15: $0.32 / $1.28 input / output, per 1M tokens | No reliable public source | No reliable public source |
| Qwen3.6-35B-A3B Qwen3.6-35B-A3B is an open-weight Qwen3.6 model available for local deployment and API use. | Qwen3.6-35B-A3B | 2026-04-16 | OpenRouter listed | OpenRouter · 2026-07-15: $0.14 / $1 input / output, per 1M tokens | No reliable public source | 1 community discussion / 1 activity snapshot |
| Qwen3.6-27B Qwen3.6-27B is an open-weight Qwen3.6 model available for local deployment and API use. | Qwen3.6-27B | 2026-04-22 | OpenRouter listed | OpenRouter · 2026-07-15: $0.45 / $2.7 input / output, per 1M tokens | No reliable public source | 1 activity snapshot |
| Qwen3.5-Omni-Plus Qwen3.5-Omni-Plus provides unified understanding and interaction across text, images, video, and audio. | qwen3.5-omni-plus | 2026-03-30 | Not separately listed on OpenRouter | Official pricing page | No reliable public source | No reliable public source |
| Qwen3.5-Omni-Flash Qwen3.5-Omni-Flash provides unified understanding and interaction across text, images, video, and audio. | qwen3.5-omni-flash | 2026-03-30 | Not separately listed on OpenRouter | Official pricing page | No reliable public source | No reliable public source |
| Wan 2.7 Image Pro Wan 2.7's high-quality image generation and editing API. | wan2.7-image-pro | 2026-04-01 | Not separately listed on OpenRouter | Official pricing page | No reliable public source | No reliable public source |
| Wan 2.7 Text-to-Video The current Wan 2.7 text-to-video generation model line. | wan2.7-t2v-2026-06-12 | 2026-06-12 | Not separately listed on OpenRouter | Official pricing page | No reliable public source | No reliable public source |
| Wan 2.7 Reference-to-Video Wan 2.7 reference-to-video API with subject referencing, voice customization, and scripted video generation from a multi-panel storyboard image. | wan2.7-r2v-2026-06-12 | 2026-06-12 | Not separately listed on OpenRouter | Official pricing page | No reliable public source | No reliable public source |
| CosyVoice 3.5 Plus Alibaba Cloud's current high-quality speech synthesis and voice-control model. | cosyvoice-v3.5-plus | 2026-03-02 | Not separately listed on OpenRouter | Official pricing page | No reliable public source | No reliable public source |
| Qwen3.5-OCR Qwen3.5 OCR model for extracting text and structure from documents and images, currently China-region only. | qwen3.5-ocr | 2026-06-16 | Not separately listed on OpenRouter | Official pricing page | No reliable public source | No reliable public source |
| Qwen3.6 Plus Alibaba Cloud Model Studio's current flagship Qwen3.6 Plus long-context API model. | qwen3.6-plus | Not disclosed | OpenRouter listed | OpenRouter · 2026-07-15: $0.325 / $1.95 input / output, per 1M tokens | No reliable public source | No reliable public source |
| Qwen3.6 Flash Alibaba Cloud Model Studio's current high-throughput Qwen3.6 Flash API model. | qwen3.6-flash | Not disclosed | OpenRouter listed | OpenRouter · 2026-07-15: $0.1875 / $1.125 input / output, per 1M tokens | No reliable public source | No reliable public source |
| Qwen3.6 Max Preview Alibaba Cloud Model Studio's current Qwen3.6 Max Preview reasoning-focused preview model. | qwen3.6-max-preview | Not disclosed | OpenRouter listed | OpenRouter · 2026-07-15: $1.04 / $6.24 input / output, per 1M tokens | No reliable public source | No reliable public source |
| Qwen3 Omni Flash Alibaba Cloud Model Studio's current Qwen3 Omni Flash omni-modal API model. | qwen3-omni-flash | Not disclosed | Not separately listed on OpenRouter | Official pricing page | No reliable public source | No reliable public source |
| Qwen3 Omni 30B A3B Qwen's current public Qwen3 Omni 30B A3B open-weight omni-modal family. | Qwen3-Omni-30B-A3B-Instruct | Not disclosed | Not separately listed on OpenRouter | No reliable public source | No reliable public source | 1 runtime / access record / 1 community discussion / 2 activity snapshots |
| Qwen3 235B A22B Qwen's current public Qwen3 235B A22B flagship open-weight MoE model. | Qwen3-235B-A22B | Not disclosed | OpenRouter listed | OpenRouter · 2026-07-15: $0.455 / $1.82 input / output, per 1M tokens | No reliable public source | 1 community discussion / 1 activity snapshot |
| Qwen3 Coder 480B A35B Instruct Qwen's current public Qwen3 Coder 480B A35B Instruct open-weight coding model. | Qwen3-Coder-480B-A35B-Instruct | Not disclosed | OpenRouter listed | OpenRouter · 2026-07-15: $0.3 / $1 input / output, per 1M tokens | No reliable public source | 1 activity snapshot |
| Qwen3 VL 235B A22B Qwen's current public Qwen3 VL 235B A22B flagship multimodal open-weight family. | Qwen3-VL-235B-A22B-Instruct | Not disclosed | OpenRouter listed | OpenRouter · 2026-07-15: $0.21 / $1.9 input / output, per 1M tokens | No reliable public source | 1 runtime / access record / 1 community discussion / 2 activity snapshots |
| Qwen3Guard Gen 8B Qwen's current public Qwen3Guard Gen 8B open-weight safety and moderation model. | Qwen3Guard-Gen-8B | Not disclosed | Not separately listed on OpenRouter | No reliable public source | No reliable public source | 1 activity snapshot |
| Qwen3Guard Stream 8B Qwen's current public Qwen3Guard Stream 8B open-weight streaming safety and moderation model. | Qwen3Guard-Stream-8B | Not disclosed | Not separately listed on OpenRouter | No reliable public source | No reliable public source | 1 community discussion / 1 activity snapshot |
| Qwen3 TTS Alibaba Cloud Model Studio's current Qwen3 TTS speech-synthesis API family. | qwen3-tts-flash | Not disclosed | Not separately listed on OpenRouter | Official pricing page | No reliable public source | No reliable public source |
| Qwen3 TTS Realtime Alibaba Cloud Model Studio's current Qwen3 realtime speech-synthesis API family. | qwen3-tts-flash-realtime | Not disclosed | Not separately listed on OpenRouter | Official pricing page | No reliable public source | No reliable public source |
| Qwen-Audio 3.0 Realtime Plus Qwen-Audio 3.0 Realtime Plus is a real-time speech model for speech reasoning and duplex conversation, with parallel inference and omnidirectional streaming used to control end-to-end response latency. | qwen-audio-3.0-realtime-plus | 2026-07-14 | Not separately listed on OpenRouter | No reliable public source | No reliable public source | No reliable public source |
| Qwen-Audio 3.0 Realtime Flash Qwen-Audio 3.0 Realtime Flash is a real-time speech model for speech reasoning and duplex conversation, with parallel inference and omnidirectional streaming used to control end-to-end response latency. | qwen-audio-3.0-realtime-flash | 2026-07-14 | Not separately listed on OpenRouter | No reliable public source | No reliable public source | No reliable public source |
| Qwen-Audio 3.0 TTS Plus Qwen-Audio 3.0 TTS Plus is a speech-synthesis model for high-quality professional scenarios, with expanded minority-language and Chinese-dialect support plus enhanced instruction following and fine-grained tag control. | qwen-audio-3.0-tts-plus | 2026-07-14 | Not separately listed on OpenRouter | No reliable public source | No reliable public source | No reliable public source |
| Qwen-Audio 3.0 TTS Flash Qwen-Audio 3.0 TTS Flash is a speech-synthesis model for low-latency real-time interaction, with time to first audio under 200 ms plus expanded minority-language and Chinese-dialect support, enhanced instruction following, and fine-grained tag control. | qwen-audio-3.0-tts-flash | 2026-07-14 | Not separately listed on OpenRouter | No reliable public source | No reliable public source | No reliable public source |
| CosyVoice Alibaba Cloud Model Studio's current CosyVoice speech-synthesis API family. | cosyvoice-v3-flash | Not disclosed | Not separately listed on OpenRouter | Official pricing page | No reliable public source | No reliable public source |
| Wan 2.6 Image Alibaba Cloud Model Studio's current Wan 2.6 image-generation and editing API model. | wan2.6-image | Not disclosed | Not separately listed on OpenRouter | Official pricing page | No reliable public source | No reliable public source |
| Wan 2.6 Image-to-Video Alibaba Cloud Model Studio's current Wan 2.6 image-to-video API model. | wan2.6-i2v | Not disclosed | Not separately listed on OpenRouter | Official pricing page | No reliable public source | No reliable public source |
| Qwen3 Embedding 8B Qwen's current public Qwen3 Embedding 8B text embedding model. | Qwen3-Embedding-8B | Not disclosed | Not separately listed on OpenRouter | No reliable public source | No reliable public source | 1 activity snapshot |
| Qwen3 Reranker 8B Qwen's current public Qwen3 Reranker 8B text reranking model. | Qwen3-Reranker-8B | Not disclosed | Not separately listed on OpenRouter | No reliable public source | No reliable public source | 1 community discussion / 1 activity snapshot |
| Qwen3 VL Embedding 8B Qwen's current public Qwen3 VL Embedding 8B multimodal embedding model. | Qwen3-VL-Embedding-8B | Not disclosed | Not separately listed on OpenRouter | No reliable public source | No reliable public source | 1 community discussion / 1 activity snapshot |
| Qwen3 VL Reranker 8B Qwen's current public Qwen3 VL Reranker 8B multimodal reranking model. | Qwen3-VL-Reranker-8B | Not disclosed | Not separately listed on OpenRouter | No reliable public source | No reliable public source | 1 runtime / access record / 1 activity snapshot |
| Qwen Image Qwen's current public Qwen Image open-weight image-generation model. | Qwen-Image | Not disclosed | Not separately listed on OpenRouter | No reliable public source | No reliable public source | 1 runtime / access record / 1 activity snapshot |
| Qwen Image 2.0 Pro Alibaba Cloud Model Studio's current Qwen Image 2.0 Pro image-generation and editing API model. | qwen-image-2.0-pro-2026-06-22 | 2026-06-25 | Not separately listed on OpenRouter | Official pricing page | No reliable public source | No reliable public source |
OpenRouter
Availability, pricing, and context
| Model | Version | Status | Pricing | Context | Time |
|---|---|---|---|---|---|
| Qwen3 235B A22B | Qwen3-235B-A22B | OpenRouter listed | $0.455 / $1.82 | 131072 | 07/15/2026, 08:11 PM UTC |
| Qwen3 Coder 480B A35B Instruct | Qwen3-Coder-480B-A35B-Instruct | OpenRouter listed | $0.3 / $1 | 1048576 | 07/15/2026, 08:11 PM UTC |
| Qwen3 VL 235B A22B | Qwen3-VL-235B-A22B-Instruct | OpenRouter listed | $0.21 / $1.9 | 131072 | 07/15/2026, 08:11 PM UTC |
| Qwen3 VL 235B A22B | Qwen3-VL-235B-A22B-Thinking | OpenRouter listed | $0.26 / $2.6 | 131072 | 07/15/2026, 08:11 PM UTC |
| Qwen3.6-27B | Qwen3.6-27B | OpenRouter listed | $0.45 / $2.7 | 262144 | 07/15/2026, 08:11 PM UTC |
| Qwen3.6-35B-A3B | Qwen3.6-35B-A3B | OpenRouter listed | $0.14 / $1 | 262144 | 07/15/2026, 08:11 PM UTC |
| Qwen3.6 Flash | qwen3.6-flash | OpenRouter listed | $0.1875 / $1.125 | 1000000 | 07/15/2026, 08:11 PM UTC |
| Qwen3.6 Max Preview | qwen3.6-max-preview | OpenRouter listed | $1.04 / $6.24 | 262144 | 07/15/2026, 08:11 PM UTC |
| Qwen3.6 Plus | qwen3.6-plus | OpenRouter listed | $0.325 / $1.95 | 1000000 | 07/15/2026, 08:11 PM UTC |
| Qwen3.7-Max | qwen3.7-max | OpenRouter listed | $1.475 / $4.425 | 1000000 | 07/15/2026, 08:11 PM UTC |
| Qwen3.7-Plus | qwen3.7-plus | OpenRouter listed | $0.32 / $1.28 | 1000000 | 07/15/2026, 08:11 PM UTC |
| CosyVoice | cosyvoice-v3-flash | Not separately listed on OpenRouter | Not applicable | Not disclosed | 07/15/2026, 08:11 PM UTC |
Hot Ranking
Linked ranking rows
| Source | Rank | Model | Value | Time |
|---|---|---|---|---|
| OpenRouter weekly token usage | 39 | Qwen: Qwen3.7 Plus | 162,784,895,917 tokens Current snapshot | 07/15/2026, 08:16 PM UTC |
| OpenRouter weekly token usage | 48 | Qwen: Qwen3.7 Max | 137,242,547,016 tokens Current snapshot | 07/15/2026, 08:16 PM UTC |
Public activity counters
Cumulative Hugging Face model-card counters
| Source | Model | Count | Link |
|---|---|---|---|
| huggingface | Qwen3.6-35B-A3B | 6,713,488 downloads / 2,397 likes | View source |
| huggingface | Qwen3.6-27B | 5,079,185 downloads / 1,976 likes | View source |
| huggingface | Qwen3 Embedding 8B | 2,887,306 downloads / 745 likes | View source |
| huggingface | Qwen3 Omni 30B A3B | 1,807,711 downloads / 957 likes | View source |
| huggingface | Qwen3 VL 235B A22B | 1,720,887 downloads / 407 likes | View source |
| huggingface | Qwen3 VL Embedding 8B | 1,478,691 downloads / 456 likes | View source |
| huggingface | Qwen3 Reranker 8B | 965,224 downloads / 248 likes | View source |
| huggingface | Qwen3 235B A22B | 930,487 downloads / 1,100 likes | View source |
Public signals
Public activity and review entry points
| Source | Target | Type | Value | Link |
|---|---|---|---|---|
| huggingface | Qwen Image | Public activity | 189552 downloads / 2538 likes | View source |
| huggingface | Qwen3 235B A22B | Public activity | 930487 downloads / 1100 likes | View source |
| huggingface | Qwen3.6-27B | Public activity | 5079185 downloads / 1976 likes | View source |
| huggingface | Qwen3.6-35B-A3B | Public activity | 6713488 downloads / 2397 likes | View source |
| huggingface | Qwen3 Coder 480B A35B Instruct | Public activity | 29972 downloads / 1353 likes | View source |
| huggingface | Qwen3 Embedding 8B | Public activity | 2887306 downloads / 745 likes | View source |
| huggingface | Qwen3 Omni 30B A3B | Public activity | 1807711 downloads / 957 likes | View source |
| huggingface | Qwen3 Omni 30B A3B | Public activity | 362765 downloads / 311 likes | View source |
| huggingface | Qwen3 Reranker 8B | Public activity | 965224 downloads / 248 likes | View source |
| huggingface | Qwen3 VL 235B A22B | Public activity | 1720887 downloads / 407 likes | View source |
| huggingface | Qwen3 VL 235B A22B | Public activity | 13381 downloads / 400 likes | View source |
| huggingface | Qwen3 VL Embedding 8B | Public activity | 1478691 downloads / 456 likes | View source |
| huggingface | Qwen3 VL Reranker 8B | Public activity | 219241 downloads / 156 likes | View source |
| huggingface | Qwen3Guard Gen 8B | Public activity | 52158 downloads / 123 likes | View source |
| huggingface | Qwen3Guard Stream 8B | Public activity | 2271 downloads / 38 likes | View source |
| huggingface | Qwen3.6-35B-A3B | Public model discussion | IDEA: Bitnet 1.58 (a4.8) version in future variants would be so incredible! | View source |
| huggingface | Qwen3 Omni 30B A3B | Model runtime/access signal | Severe generation length bottleneck (hardcoded at 1024 tokens) in Qwen3OmniMoeForConditionalGeneration during greedy search / custom generation loop | View source |
| huggingface | Qwen3Guard models below 8B | Public model discussion | Models below 8B cant detect simple jailbreak prompts | View source |
| huggingface | Qwen3 Omni 30B A3B | Public model discussion | Smaller version such as 8b or 4b will it be released a dn when? | View source |
| huggingface | Qwen3 VL Reranker 8B | Model runtime/access signal | Today's template update breaks sglang | View source |
| huggingface | Qwen3 VL 235B A22B | Public model discussion | Avoid Re-encoding Reference Images in Vision-LLM When Comparison Criteria Are User-Defined | View source |
| Qwen3.5-35B-A3B | Public snapshot discussion | Qwen3.5 feels ready for production use | View source | |
| huggingface | Qwen3 VL Embedding 8B | Public model discussion | GGUF Quantization | View source |
| huggingface | Qwen3 Reranker 8B | Public model discussion | Working GGUF for llama.cpp (native Windows/Linux, no WSL needed) | View source |
| Qwen3-Coder-Next | Public snapshot discussion | Qwen Coder Next is an odd model | View source | |
| huggingface | Qwen3 VL 235B A22B | Model runtime/access signal | Qwen3-VL-235B-A22B-Instruct-AWQ returns chinese characters | View source |
| huggingface | Qwen Image | Model runtime/access signal | Concurrent requests on one Qwen-Image pipeline trigger an index-out-of-bounds error | View source |
| huggingface | Qwen3 235B A22B | Public model discussion | What is the minimum hardware needed to run Qwen3-235B-A22B with vLLM? | View source |
| huggingface | Qwen3Guard Stream 8B | Public model discussion | How can I tune the model output towards business needs? Can I do fast prompt tuning? | View source |
| Qwen3 family | Public snapshot discussion | New Qwen models are unbearable | View source |
Evidence gaps
Company-level display boundary
Public-signal coverage is No safe source