OpenAI Supercharges Sora Video Generation Tool

OpenAI just rolled out significant upgrades to its Sora video generation API, giving creators powerful new tools for professional-quality content production. Based on the improved Sora 2 model, these updates solve some of the most common pain points in batch video creation.

Characters That Stay Consistent

The standout feature addresses what many users found frustrating - characters that kept changing appearance between scenes. Previously, generating multiple clips often resulted in subtle (and sometimes not-so-subtle) shifts in facial features, clothing, and accessories.

Now creators can upload detailed "character profiles" defining appearance elements. The system maintains these references across all generated segments, ensuring visual continuity perfect for serial content like ads or web series.

Longer Videos With Smoother Flow

The update extends maximum clip length from 16 seconds to a more practical 20 seconds - enough time for complete narrative beats without awkward stitching. An innovative extension feature lets the AI naturally continue existing clips too, opening possibilities for longer-form storytelling.

Ready-to-Share Formats

In today's multi-platform world, the new dual-output capability saves hours of work. A single generation now produces both:

• 16:9 landscape versions ideal for YouTube and desktop viewing
• 9:16 portrait formats optimized for TikTok and mobile feeds No more manual cropping or quality loss from reformatting.

Behind-the-Scenes Improvements

The update also boosts batch processing power through enhanced asynchronous support - great news for studios managing large rendering queues or automated pipelines.

Key Points:

• Character consistency maintained across multiple scenes
• Extended duration up to 20 seconds per clip
• Dual output formats eliminate manual reformatting
• Batch processing enhancements streamline workflows
• Video extension allows natural continuation of clips