Sora v2 is coming with longer clips and multimodality
OpenAI is gearing up to launch Sora v2 very soon, according to creative director Chad Nelson at a London keynote. The upgraded model can generate videos up to one minute long, adding image-plus-text and video-to-video capabilities alongside standard text prompts.
OpenAI's Chad Nelson just dropped some major hints about what is coming next during a keynote at the C21Media event in London.
According to an attendee, the launch of Sora v2 is now impending. Sam Altman had already teased it, and Nelson confirmed on stage that people will see it very soon.
Longer clips and flexible inputs
The upcoming update seems directly aimed at making the model more practical for actual production workflows.
Sora v2 will be capable of generating one-minute video outputs.
Just as importantly, it expands beyond standard text-to-video prompts. The new model supports text-plus-image inputs as well as text-plus-video inputs, allowing creators to guide the generation using existing visual assets.
Why it matters
Text prompts alone are notoriously tricky for video creators who need exact character consistency or specific camera movements. Allowing images and existing video clips as references gives editors real control instead of forcing them to roll the dice.
In our view, OpenAI needs this upgrade fast. Competitors have sprinted ahead by offering flexible image-to-video and clip editing tools. If Sora v2 can deliver one-minute coherence while letting creators steer with their own visual inputs, it looks ready to graduate from a jaw-dropping viral demo into a dependable creative tool.
Sources (1)
- @RuudNL(opens in a new tab)Originaltwitter.com