Video
GA
Open source
Open-Sora 2.0
Video research with a fully open dataset and training pipeline
Release2025-03
Context4K
Max output—
Parameters11B
Local / VRAM24–80 GB
LicenseOpen source
Overview
Open-Sora 2.0 is a model in the Video category from HPC-AI Tech (Open-Sora). It was released in 2025-03. The vendor is based in Singapore. Best for: Video research with a fully open dataset and training pipeline.
Strengths
- Apache-2.0
- Open training code
- 16 sec
Limits
- Not quality-leading
- No audio
Specifications
| Vendor | HPC-AI Tech (Open-Sora) |
|---|---|
| Category | Video |
| Release | 2025-03 |
| Status | GA |
| License | Open source |
| License name | Apache-2.0 |
| Commercial use | Allowed |
| Parameters | 11B |
| Architecture | DiT |
| Context window | 4K |
| Input modalities | Text, Image |
| Output modalities | Video |
| Prompt types | Text |
| Negative prompts | No |
| Max outputs | 1 clip |
| Resolution | 480p-720p |
| Duration | 5-16s |
| Frame rate | 24 fps |
| Audio generation | No |
| Local deployment | Yes |
| VRAM | 24–80 GB |
| System RAM | 32–128 GB |
| API access | Yes |
| Pricing | Free weights |
| Confidence | High |
| Vendor country | Singapore |
| Vendor type | Research |
Pricing & hardware
Free weights
This model can be self-hosted or run locally. Approximate GPU memory: 24–80 GB. Approximate system RAM: 32–128 GB. API access is available.