turn a single image into a speaking video image-to-video is now live for everyone powered by sync-3

RT @naval: Introducing USVC - a single basket of high-growth venture capital, for everyone. No accreditation required, SEC-registered, an…

not all shots are created equal its easy to lipsync someone when they are facing the camera directly, but what happens when they are not? when they are looking to the side, or when they are putting on lipstick? what about when they are speaking in a mirror? sync-3 handles all that and more with ease, giving you the ability to tell your stories however you wish

today we are launching sync-3: the next evolution in lipsync every model we've ever built has led to this every story for every audience

every model we've ever built has led to this an evolution in lipsync sync-3 coming soon

lipsync, now in Claude Code and Codex introducing the sync. MCP server generate lipsynced videos, manage assets, and track renders - all from your code. works with Codex, Claude Code, VS Code, Cursor, and any MCP client vibe-editing just got real -> go.sync.so/MCP

sync. 🤝 premiere pro the magic of sync. meets the power of premiere pro edit, lipsync, dub, generate - all in one tool never leave your timeline again

editing a conversation should be as easy as having one. today we are excited to announce the launch of the sync. assistant - a whole new way to lipsync your videos

We can talk all day about how these models are going to produce so much good and produce value. And I think they will. But the reality is many of my buddies are unemployed, can't seem to find jobs, and continue to send 1,000 resumes a week on LinkedIn. This is even worse for new grads. These models are wonderful tools but a majority of people don't know how to do much other than write emails with them. Just because you invent a tool like the hammer, doesn't mean everyone is going to magically know how to build a house with it. I don't think people are lazy. Last week I made a tweet doing a free Codex class and 5,000 people came by. It's clear people want to build and adapt, they just don't know where to start. The appetite is there, but nearly all of the scaffolding education is missing. AI people love to quote the industrial revolution all the time, and how back then people were scared of machines taking their jobs but it all worked out. What they never mention is that it only worked out because we built an entire education system from scratch to bring everyone along. The industrial revolution created a massive wave of new jobs and nobody was trained for any of them. The tools didn't save anyone on their own, because the education had to catch up. It's never been a better time to start an education company. Millions of people want to learn, many of the tools are free, and the only thing missing is someone showing them the way.

Cinema has no language barrier 🎬✨

hollywood is decentralizing, and its incredible to watch

Merry Christmas from our family to yours🎄 this holiday season, gift your loved ones the joy of creativity!

today, sync. is excited to announce react-1 a ten billion parameter masked video diffusion model built for one purpose: giving you full creative control over performances in post

react-1 is the first performance control primitive for video editing once you film a shot, a character's performance is locked – you can't iterate without risking changing the video entirely, or spending weeks in vfx tools react-1 changes this. record a scene, then fine-tune the performance in post. explore different emotional reads, adjust timing, refine the acting — all after the camera turns off the workflow: film → direct → refine

"acting is reacting" react-1, coming soon to a screen near you

react-1 allows you to modify only the region you want, leaving the rest of your video untouched game developers, animators, and filmmakers - this tool is for you elevate your craft by controlling even the smallest details

with react-1, you don’t just pick the emotion for your character you control exactly how intense it feels giving you the creative control your story demands

become the director you’ve always wanted to be react-1 lets you edit any scene, allowing you to shape the performance, timing, and expression coming soon

traditional AI models just move the mouth to match the audio. react-1 changes this entirely. it uses the audio to rebuild the whole performance: timing, emotion, and micro-expressions. instead of just looking dubbed, the actor looks like they meant every word in that language.

until now, AI videos were permanent the second they finished generating react-1 changes that you can now edit a finished clip: expressions, tone, lipsync, head direction, presence, all after generation here is a preview of what's in store