All tasks / Models
daVinci-MagiHuman
Open-weights 15B single-stream model that generates talking-human video with synchronized speech from text and an image.
- SII-GAIR and Sand.ai
- Open weights
- Released Mar 2026
Where it ranks
- –Video generationModels · Local · no result on this task’s benchmarks yet
- –Avatars & dubbingModels · Local · no result on this task’s benchmarks yet
Facts
- Vendor
- SII-GAIR and Sand.ai
- Family
- daVinci-MagiHuman
- Released
- Mar 2026
- Weights
- Open: download and run it yourself
- License
- Apache-2.0
- Parameters
- 15B
- Access
- Free (open weights)
- Last verified
- 19 Oct 2026
Price from github.com, 19 Oct 2026.
Benchmark results
No public benchmark results recorded yet.
Where you can use it
No tracked tool lists this model yet; as open weights it runs in local runtimes.