Qwen is Alibaba’s open-weight model family. The Qwen series leads many open leaderboards for code, math, and multilingual benchmarks and includes Mixture-of-Experts and dense variants from 0.5B to 235B parameters.
See all models from AlibabaModels in family
32
Open weight
22
API only
10
Avg score
61.9
Top benchmark
94.1
AIME 2026
Total HF downloads
43.0M
Primary modality
Text
Context window
131K – 1.0M
First release
Jan 2025
Latest release
Sep 2026
Every release in the Qwen family, ranked by composite score across benchmarks, popularity, efficiency, and versatility.
| # | Model | Modality | Score | Params | Released |
|---|---|---|---|---|---|
| 1 | audio | AA79.0 | — | Jan 2026 | |
| 2 | text | AA75.2 | 35B | Apr 2026 | |
| 3 | embedding | AA74.7 | — | Jun 2025 | |
| 4 | Qwen3.5 FlashAPI | text | AA74.6 | 35B | Feb 2026 |
| 5 | text | AA72.9 | 35B | Feb 2026 | |
| 6 | embedding | AA72.5 | 7.6B | Jun 2025 | |
| 7 | audio | AA72.4 | 1.7B | Jan 2026 | |
| 8 | embedding | AA71.9 | 4B | Jun 2025 | |
| 9 | Qwen3.5 Max PreviewAPI | text | AA71.9 | — | Feb 2026 |
| 10 | text | BB69.9 | 27B | Apr 2026 |
When a family spans multiple modalities, the strongest model in each one.
How the family splits across text, image, video, audio, embedding, and decision releases.
When each release shipped, newest first. Useful for tracking version cadence.
Jun 1
May 20
Apr 21
Apr 17
Apr 13
Mar 31
Mar 28
Mar 1
Feb 23
Feb 23
Feb 23
Feb 23
Feb 15
Feb 15
Jan 29
Jan 29
Dec 1
Nov 1
Sep 22
Sep 22
Jun 5
Jun 5
Jun 5
Apr 27
Apr 27
Apr 27
Jan 24
Jan 24
Composite grades across this family. Higher is better, blending benchmarks, popularity, and efficiency.
Models with downloadable weights, ranked by composite score.
| # | Model | Modality | Score | Params | Released |
|---|---|---|---|---|---|
| 1 | audio | AA79.0 | — | Jan 2026 | |
| 2 | text | AA75.2 | 35B | Apr 2026 | |
| 3 | embedding | AA74.7 | — | Jun 2025 | |
| 4 | text | AA72.9 | 35B | Feb 2026 | |
| 5 | embedding | AA72.5 | 7.6B | Jun 2025 | |
| 6 | audio | AA72.4 | 1.7B | Jan 2026 | |
| 7 | embedding | AA71.9 | 4B | Jun 2025 | |
| 8 | text | BB69.9 | 27B | Apr 2026 | |
| 9 | text | BB68.6 | 397B | Feb 2026 | |
| 10 | text | BB68.2 | 27B | Feb 2026 | |
| 11 | image | BB67.1 | 20B | Nov 2025 | |
| 12 | text | BB65.4 | 397B | Mar 2026 | |
| 13 | text | BB63.2 | 9B | Mar 2026 | |
| 14 | text | BB62.3 | 122B | Feb 2026 | |
| 15 | text | BB58.6 | 30B | Apr 2025 | |
| 16 | image | CC53.7 | 20B | — | |
| 17 | text | CC53.3 | 235B | Apr 2025 | |
| 18 | text | CC52.9 | 32.8B | Apr 2025 | |
| 19 | image | CC51.8 | 20B | Dec 2025 | |
| 20 | image | CC42.1 | 20B | — | |
| 21 | decision | DD31.8 | — | Sep 2026 | |
| 22 | decision | DD29.6 | 9.7B | Sep 2026 |
Closed-source releases accessed through Alibaba’s API.
| # | Model | Modality | Score | Params | Released |
|---|---|---|---|---|---|
| 1 | Qwen3.5 FlashAPI | text | AA74.6 | 35B | Feb 2026 |
| 2 | Qwen3.5 Max PreviewAPI | text | AA71.9 | — | Feb 2026 |
| 3 | Qwen3.6-PlusAPI | text | BB68.6 | — | Mar 2026 |
| 4 | Qwen3.6 Max PreviewAPI | text | BB68.0 | — | Apr 2026 |
| 5 | Qwen 3.7 MaxAPI | text | BB66.3 | — | May 2026 |
| 6 | Qwen Plus (0125)API | text | BB62.0 | — | Jan 2025 |
| 7 | Qwen 3.7 PlusAPI | text | BB57.8 | — | Jun 2026 |
| 8 | Qwen3 Max (2025-09-23)API | text | CC53.2 | — | Sep 2025 |
| 9 | Qwen3 Max PreviewAPI | text | CC50.1 | — | Sep 2025 |
| 10 | Qwen2.5 MaxAPI | text | CC49.7 | — | Jan 2025 |
The Qwen family is a series of AI models from Alibaba. This page lists every release in the family with its benchmark scores, parameter count, and hardware requirements.
By composite score, Qwen3-ASR-0.6B is currently the top model in the family. For local inference, match the parameter count to your VRAM budget. For quality, pick the highest scorer that fits.
See the open-weight section above for models you can run locally. The API-only section lists closed releases that must be accessed through the provider’s API.
Spin up an instance in the cloud, or pick local hardware that fits.
Advertising disclosure: we earn commissions when you shop through the links below.
Vast.ai
Decentralized GPU marketplace with the lowest hourly prices.
RunPod
Pay-per-second GPU rentals starting at $0.20/hr.
Digital Ocean
Spin up a GPU droplet in minutes, starting at $0.75/hr.
Vultr
GPU cloud with hourly and monthly plans, starting at $0.50/hr.
GPU Mart
Dedicated GPU servers and VPS billed monthly, starting at $0.50/hr.
PPQ.ai
Multi-model inference gateway for production workloads.
Find Local Hardware
See which GPU, Mac, or workstation can run Qwen on-prem.
Local LLM Mini PCs
Compact machines that run Qwen on your desk. Chosen for local inference.