Open-Sora
An open-source video generation model for text-to-video, with a maximum duration of 16 seconds, driven by the community.
📋 Info
| GitHub Stars | ⭐ 29.1k Stars |
| License | Apache-2.0 |
| Language | Python |
| Version | 2.0.0 |
| Updated | 2026-04-20 |
📖 Overview
Open-Sora is an open-source project created by the community to replicate OpenAI Sora, and it is currently one of the most active open-source video generation solutions (with 29k Stars). It supports text-to-video generation and is built on an efficient Diffusion Transformer architecture. The community continuously works to improve model quality and output length, representing the cutting-edge advancements in open-source video generation. It is suitable for developers and researchers interested in the latest technologies in AI video generation.
✨ Features
- Text-to-video generation (up to 16 seconds long)
- Diffusion Transformer architecture
- Ongoing optimization driven by the community.
- Pre-trained model + inference script
- Open-source cutting-edge text-to-video capabilities
Advertisement