Hayz Labs · https://hayzlabs.com/reviews/makers/alibaba/
makers · Alibaba
Alibaba
Everything from Alibaba we track, and our call on each. Its strongest is Wan 3.0 for video.
- Scored
- 2
- First looks
- 0
- On the list
- 9
- Passed on
- 0
Reviewed
Best score first; each shows its best job. First looks are new and not scored yet.
| Name | Score | Cheapest plan | Our verdict | Why |
|---|---|---|---|---|
| 89Video | Pay as you go | Use it | Top-tier video from text or images, clips up to 30 seconds, at half Veo 3.1's price. MiniMax H3 costs less. | |
Qwen-Image-2.1Open source · best at images | 48Images | Free | Depends | Best scoring image model you can download and run yourself, but its license bars commercial use. Fine for research. |
On the list, not reviewed
Big news or good value, on our Top 20 lists.
- Qwen3.8-MaxAlibaba's hosted flagship; scores near the top Western models on web app coding and vision at a fraction of their price.
- Qwen3.8-27BBest open model that fits one high-end GPU: Apache-2.0, image and video input, 262K context, and the most fine-tuned base on Hugging Face.
- Qwen-Audio-3.1-TTS-PlusRanks #3 in blind voice tests, about 30 Elo behind ElevenLabs' best, for under half the price per character.
- Qwen-Image-3.0Strong at dense text, infographics and Chinese characters for 3 to 4 cents; unlike earlier Qwen-Image, no open weights yet.
- Qwen Audio 3.0 RealtimeSecond-best realtime voice model on Artificial Analysis at a small fraction of GPT-Realtime's audio price.
- Qwen3-Embedding-8BOpen embedding model near the top of the MTEB leaderboard that you can run yourself.
- Qwen3.7-PlusLow-cost Qwen model for simple everyday tasks
- Fun-ASR-FlashLowest error rate on the board; for teams able to use Alibaba Cloud Model Studio
- Qwen3.5 Omni Plus RealtimeTalks and also sees video; behind the leaders in blind votes, mainly for Alibaba Cloud users