Adobe Launches Generative Extend AI Tool in Premiere Pro 25.2

Generate those extra frames instead of reshooting
Generative Extend lets editors add up to two seconds of AI-generated footage to smooth over unexpected moments.
Mark

So Generative Extend can add two seconds to a video clip. That sounds small. Why does that matter?

Mimi

Two seconds is actually significant in editing. You might have a shot where someone moves unexpectedly, or you need a bit more time for a transition to land. Instead of reshooting or finding a different clip, you can generate those extra frames.

Luke

But it's generating them, right? So the quality—is that proven? Adobe says 4K, but we don't know how convincing the generated frames actually look in real editing scenarios.

Mimi

Fair point. Adobe hasn't shown extensive before-and-after examples yet. What we know is it's built on Firefly, which is their video model, and it works in 4K. The real test will be when editors start using it.

Mark

And the audio side—it extends ambient sound up to ten seconds but won't touch speech or music. Why that limitation?

Mimi

Probably because extending speech or music would require understanding content and meaning in ways the AI isn't ready for. Ambient sound is more forgiving—wind, room tone, that kind of thing.

Luke

Right, but that also means the tool is narrower than it sounds. You can't use it to extend a voiceover or a music bed, which are things editors actually need.

Mark

When does Adobe start charging?

Mimi

They haven't said exactly. It's free now, but they've committed to charging Firefly credits eventually. Pricing will vary by format, frame rate, and resolution.

Luke

Which means we don't know the actual cost yet. That's important context—editors might adopt it now and face sticker shock later.

Mark

What about the search tool? That sounds genuinely useful.

Mimi

It is. You describe what you're looking for—a wide shot, a close-up of a face, a specific location—and it finds it in your footage. That's real time saved.

Luke

Again, though, we're trusting the AI to understand your description correctly. If it misses things or returns false positives, it could slow you down instead of speeding you up.

  • Video editors have long been at the mercy of imperfect footage, but Generative Extend now lets them stretch clips by up to two seconds in 4K, turning near-misses into usable moments.
  • The tool's free availability creates immediate urgency — editors who adopt it now will eventually face a pricing shift tied to format, frame rate, and resolution, making the clock a quiet pressure in the background.
  • AI-powered text search, automatic caption translation across 27 languages, and log-to-HDR color conversion are compressing what once took hours of manual labor into a matter of keystrokes.
  • After Effects 25.2 joins the push, adding 3D FBX model support, HDR monitoring, and smoother preview playback — signaling that Adobe's AI integration is a suite-wide strategy, not a single feature experiment.
  • The industry is watching to see whether editors embrace these tools as creative liberators or resist them as the first foothold of automation in a deeply human craft.

In the ongoing negotiation between human creativity and machine assistance, Adobe has quietly shifted the terms of video editing with the release of Premiere Pro 25.2. At its heart is Generative Extend, a tool built on the Firefly AI model that grants editors something time itself rarely offers: a second chance at a clip already captured. Free for now but destined for a credit-based pricing model, this update — alongside AI search, automatic caption translation, and companion upgrades to After Effects — marks another step in the slow transformation of technical craft into conversational collaboration.

Adobe has released Premiere Pro 25.2, centering the update around Generative Extend — an AI tool that allows editors to add up to two seconds of footage to an already-shot clip in 4K quality. Built on Adobe's Firefly video model, first previewed in October, the feature is designed to smooth over unwanted movements, create space for transitions, or simply recover frames that weren't quite there. It can also extend ambient background audio by up to ten seconds independently, or two seconds when paired with video extension. Music and speech tracks are deliberately left untouched. The tool is currently free, though Adobe has confirmed it will eventually require Firefly generative credits, with costs varying by format, frame rate, and resolution.

Generative Extend shares the update with several other AI-driven additions. A new search panel lets editors describe footage in plain language — by object, camera angle, location, or effect — replacing manual timeline scrubbing with something closer to a conversation. A color management tool handles the conversion of log and raw files into SDR and HDR formats without lookup tables, and arrives with three wide-gamut presets. Caption translation now covers 27 languages automatically, with multiple caption tracks viewable simultaneously during editing. Performance improvements across Apple and Windows round out the release.

Adobe also updated After Effects 25.2 in parallel, adding 3D FBX model support, animated environments, HDR monitoring, new animation controls, and a High Performance Preview Playback feature to reduce lag on longer compositions. Taken together, these releases reflect Adobe's broader ambition: to embed AI assistance so thoroughly into its video suite that technically demanding tasks become intuitive, and the remaining question is less about capability than about cost — and whether editors will find the eventual pricing model worth the creative ground gained.

Adobe has rolled out Premiere Pro 25.2, and at the center of this update sits Generative Extend, an AI tool that does something video editors have wanted for years: it lets you add time to a clip that's already been shot. The tool can stretch video footage by up to two seconds in 4K quality, which means you can smooth over unexpected movements, create room for new transitions, or simply buy yourself a few extra frames to work with. For now, it's free to use, though Adobe has made clear that won't last forever.

The Generative Extend feature runs on Adobe's Firefly video model, technology the company first previewed in October. Beyond just extending the visual content, the tool can also stretch ambient background audio up to ten seconds on its own. If you're extending both video and audio together, the limit drops to two seconds for the audio portion. The system won't touch music or speech tracks, only ambient sound—a deliberate constraint that keeps the tool focused on its core job. When Adobe does begin charging for the feature, pricing will shift based on your video's format, frame rate, and resolution.

But Generative Extend isn't traveling alone in this update. Adobe has also introduced an AI-powered search panel that lets editors hunt through footage using text descriptions. You can search by object, camera angle, location, or effects, turning what used to be a manual hunt through timelines into a conversation with the software. A new color management tool handles the technical grunt work of converting log and raw files into SDR and HDR formats without requiring lookup tables, and it comes with three wide-gamut presets ready to go.

Caption work has gotten faster too. Premiere Pro can now automatically translate video captions into 27 different languages using AI, and editors can display multiple caption tracks simultaneously during the editing process. The update also promises better speed and performance across both Apple and Windows machines, addressing one of the perpetual complaints about video editing software: that it can feel sluggish when you're working with large files.

Adobe hasn't stopped at Premiere Pro. The company also released After Effects 25.2, which adds support for 3D FBX models, animated environments and light effects, new HDR monitoring capabilities, and fresh animation controls. A new High Performance Preview Playback feature makes it easier to preview longer compositions without the usual stuttering and lag. Together, these updates represent Adobe's push to embed AI assistance throughout its video creation suite, turning tasks that once required technical knowledge or brute-force searching into something more conversational and immediate. The question now is how quickly editors will adopt these tools and whether the eventual pricing model will feel fair once the free trial period ends.

Price will vary based on the format, frame rate, and resolution of your video.
— Adobe announcement
Fale Conosco FAQ