Sora
OpenAI's frontier video model, now its own social app
Our verdict
The most impressive raw generations available — but plan tier and rate limits decide how much you can actually make.
Sora generates longer, story-coherent clips than most rivals and can remix uploaded footage, with a feed built around it. Access comes through ChatGPT plans rather than a separate subscription.
How we scored it
What we liked
- ✓Long, coherent shots
- ✓Video remixing and uploads
- ✓Free daily allowance
What to watch
- ✕Content rules block some prompts
- ✕Realistic physics still slips
- ✕Rate limits decide how much you can make
Key features
- ◆Text-to-video
- ◆Video remix
- ◆Storyboard sequences
- ◆Character consistency
- ◆Community feed
Best for
Appears in
Alternatives to Sora
Runway
The full AI video suite — generate, edit and finish in one place
Runway is more than a generator: Gen-4 text and image-to-video sits beside a browser editor with masking, motion brush, colour tools and effects. It is the closest thing to an AI-native post-production pipeline for short-form creators.
Veo
Google's frontier video model, in Gemini, Flow and Vertex
Veo is Google DeepMind's flagship video generator, accessible in Gemini, the Flow filmmaking app and Vertex AI. It is among the best at native audio, physical realism and following complex direction.
Luma Dream Machine
Fast, physical video generation with a generous free tier
Luma's Dream Machine generates smooth, physically believable motion from text or a still, and adds keyframes, loops and an editor for assembling shots. Generation is fast and the free allowance is real.
Descript
Edit video by editing the transcript
Descript treats your recording as a document: cut a word out of the transcript and it disappears from the video, then layer on studio sound, eye contact correction and filler-word removal. It is built for podcasts, interviews and talking-head YouTube.