Model
OpenAI Sora
Run OpenAI Sora 2 and Sora 2 Pro for text-to-video and image-to-video, with strong scene coherence over longer shots.
Sora is OpenAI's video model, notable for scene coherence — it holds a consistent world across a shot rather than drifting, which shows most in sequences with camera movement or several interacting subjects.
Sora 2 and Sora 2 Pro are both available, in text-to-video and image-to-video forms. Pro costs more and is the one to use for final output; the standard model is the sensible choice while you are still iterating on a prompt.
Sora carries the 1.5x premium multiplier, so a 5-second 720p render is around 30 credits, with cost scaling from there by duration and resolution.
FAQ
OpenAI Sora — questions
What is the difference between Sora 2 and Sora 2 Pro?
Pro is the higher-fidelity tier at a higher cost. Both accept text and image inputs.
Can Sora animate an image I supply?
Yes — image-to-video variants of both Sora 2 and Sora 2 Pro are in the catalog.
How does Sora compare to Veo?
Broadly, Sora holds scene coherence over longer shots while Veo is stronger on literal prompt adherence. Both are premium-tier and cost the same in credits, so it is worth trying both on the same prompt.
Keep reading
Related guides
Start generating today
100 free credits on signup — roughly fifty images, or twenty-five voiceovers, or three premium video clips.
No card required.