How does Sora generate video?

It uses large generative diffusion models to transform text or image inputs into coherent video frames over time.

Does Sora output include audio?

Yes, Sora 2 adds synced ambience, music, and basic dialogue to complete the scene.

How long can videos be?

Official limits vary by plan, with typical ranges around 10-15 seconds.

Can I control style and length?

Yes. Use prompts plus settings like aspect ratio, resolution, and duration to guide output.

AI video generation

Sora2 AI Video Generator

Turn text or images into cinematic short videos with improved physical realism, continuity, and sound. Sora 2 (2025) upgrades physical understanding, visual fidelity, and controllability for multi-shot storytelling.

Seedance.so no longer offers Sora 2 or Sora 2 Pro video generation and will not provide future support for these models.

Choose another video model

View capabilities

Why Sora 2 is no longer supported

Seedance.so no longer offers Sora 2 or Sora 2 Pro video generation and will not provide future support for these models.

OpenAI's official notices state that the Sora web and app experiences ended on April 26, 2026, and that the Videos API, sora-2 and sora-2-pro were removed on September 24, 2026. The upstream service has been retired; the linked notices do not give a more detailed business reason for its closure.

This page is retained for reference. Please choose an available video model in Media Studio for new creations.

Official OpenAI sources

Official sources checked on October 18, 2026.

Choose another video model

Key Highlights

Stronger physics consistency, more realistic results

Sora 2 follows real-world physics and cause-and-effect more closely. It avoids breaking common sense just to satisfy prompts, and instead simulates believable outcomes.

Cinematic visuals with a complete audio-visual experience

Compared with the previous version, Sora 2 greatly improves realism, lighting, texture, and motion continuity, delivering photoreal, cinematic, and high-quality animation styles. It also auto-generates matching dialogue, sound effects, and background music with tight sync for deeper immersion.

Greater control and long-range continuity

Sora 2 understands complex instructions, supports multi-shot and multi-scene generation, and keeps people, environments, and objects consistent. New “character injection” lets you bring specific faces and voices into scenes for efficient, controllable storytelling.