Sora 2
OpenAI's video model with always-on audio generation. Every Sora 2 clip comes with AI-generated sound — no toggle required. Cinematic output across multiple resolutions and aspect ratios.
What is Sora 2?
Sora 2 is OpenAI's second-generation video model. One of its defining characteristics is always-on audio — unlike models where audio is an optional toggle, Sora 2 generates AI audio as an inherent part of every clip. Every output arrives with accompanying sound.
The model produces cinematic-quality video across multiple aspect ratios including 16:9, 9:16, and 1:1, making it suited for everything from landscape cinematic content to vertical social and square formats. Multiple resolution options give you control over output fidelity.
Start frame support allows you to anchor the visual beginning of a clip with an uploaded image, directing the model's opening visual context. Combined with always-on audio and OpenAI's cinematic output quality, Sora 2 is a complete single-pass video production tool for social, advertising, and brand content.
Always-on audio
Audio in every generation
Multiple resolutions
Flexible output fidelity
Start frame
Anchor the opening visual
OpenAI quality
Cinematic output standard
How to generate video with Sora 2 on project.video
Open the composer
Go to your project.video dashboard and open the generation composer. All models including Sora 2 are available from a single unified workspace.
Select Sora 2
Click the model selector and choose Sora 2. Your prompt transfers from any other model — no re-entry needed when comparing outputs.
Write your prompt
Describe your scene with cinematic specificity. Sora 2 responds well to descriptions of camera movement, lighting, and atmosphere. Audio will be generated alongside video automatically.
Set specs and generate
Choose aspect ratio, resolution, duration, and optionally upload a start frame image. Hit generate and view your output — audio included — in your gallery.
Technical specs
Best use cases
Social video with natural soundtrack
Every Sora 2 generation includes audio — no post-production audio sourcing needed. Generate a complete social video with ambient sound, music, or dialogue-appropriate FX in a single pass.
Brand storytelling
Cinematic quality from OpenAI combined with always-on audio makes Sora 2 well-suited for narrative brand content that needs both visual and audio polish without separate production steps.
Advertising creative
Generate multiple ad creative directions — each with video and audio — in a single session. Compare visual styles and audio characters across generations to find the winning direction.
Animation from still images
Upload a start frame to animate from a specific visual — a product image, a branded illustration, or a scene setup — and let Sora 2 generate cinematic motion and audio from it.
Example prompts
Copy any of these to get started on project.video. Audio is included automatically with every Sora 2 generation.
"A jazz quartet performs in a dimly lit underground club, brass and piano filling the room, the camera slowly orbits the group, moody amber lighting, 16:9"
"Rain falls on a cobblestone street in Paris at night, reflections of café lights stretching long on the wet surface, ambient city audio, 16:9"
"A surfer paddles out into glassy morning ocean, pelicans flying above, water sounds and wind, golden sunrise behind the wave sets, vertical format, 9:16"