🚨BREAKING NEWS - LIVE FULLY AUTONOMOUS AVATAR STREAMING AND ROASTING GITHUBS ON TWITCH.🚨
HAPPENING NOW ...
Watch it now at https://t.co/LbZt7QWfxV.
BREAKING. A fully autonomous AI avatar is live on Twitch right now,
roasting GitHub repos 24 hours a day, in real time, with no script.
@HeyGen put it up and pointed it at code.
You drop a GitHub repo in the chat.
The avatar opens it, reviews the actual code, roasts the bad design decisions out loud, and navigates the browser by itself.
No human at the keyboard.
No human in the chair.
This is HeyGen's Avatar Realtime running fully autonomous.
Not a pre recorded clip, not a person behind the curtain. 🤯🤯
It reads, reacts, and drives the screen live.
Watch it now at https://t.co/LbZt7QWfxV.
We put an AI avatar on Twitch and asked it to roast 24/7
People send GitHub repos in chat and it opens them, reviews the code, roasts bad design decisions, and navigates the browser by itself in real time
https://t.co/QH7K9sP5EG
This is a full month of releases captured in one recap, and what stands out is not the count but the direction it is all pulling in.
@HeyGen is quietly turning video into something you direct rather than something you build by hand, and that shift is worth slowing down to understand.
May was packed with updates.
HyperFrames got upgrades. Avatar V now takes direction in plain English, and is available via API and Live.
We launched integrations with Superhuman, Codex, Canva, and Lovable, plus HeyGen on Android, and more.
Full recap: https://t.co/gfDVGdVCwS
Congratulations to @bfl_ai Black Forest Labs and Martin Scorsese
This is the right vision for AI, and it's the one I've been arguing for as a filmmaker: keep human taste, craft, and judgment at the center, and the tool stops replacing the artist and starts amplifying the operator.
What stands out is that he didn't hand over his vision.
He would used it to get the picture in his head to his crew faster.
The vision stayed human. That's the whole game.
Full breakdown link on this on Mediumin the comments.
Martin Scorsese is an advisor to Black Forest Labs.
He's spent six decades shaping how the world sees stories. Now he's helping us shape visual intelligence with human taste and craft at the center.
We sat down with him for a working storyboarding session using FLUX.
Congratulations to @bfl_ai Black Forest Labs and Martin Scorsese
This is the right vision for AI, and it's the one I've been arguing for as a filmmaker: keep human taste, craft, and judgment at the center, and the tool stops replacing the artist and starts amplifying the operator.
What stands out is that he didn't hand over his vision.
He used it to get the picture in his head to his crew faster. The vision stayed human.
That's the whole game.
If you want the full breakdown on this, it's on Medium: https://t.co/A0VgIVpXaa
Newsletter Video Just Became a Render and a GAME CHANGER...
This is on an another GEM @HeyGen ...
My clients are going to go bonkers for content like this...
Perfect for Newsletter Updates across so many markets...
Set the template, swap the month and numbers...
Keep your audience nurtured and repeat next week/ month/ quarter...
#gamechanger
Many financial advisors spend hundreds, sometimes thousands, per month on client update videos
We wanted to see if Avatar 5 + HyperFrames could handle the entire task
So we gave an AI agent a brief and let it run
Here's what it created, plus the workflow in the thread below ↓
Proud to be one of Inc.'s 2026 Best Workplaces.
The truth: we genuinely love the people we work with. We move fast. We care deeply about the work.
Most places make you pick one or two. We've managed to keep all three.
Thanks to every HeyGen team member who made this possible.
No crew. No actors. No country. A $2,000 A.I. film just made Tribeca.
Three months. No crew. No actors.
No access to the country it is about.
It just broke into Tribeca.
Every headline will say the AI made it.
It did not.
He did.
The tools were only the rails.
How @AshKoosha made a feature film for about $2,000.
Full breakdown:
https://t.co/BgwUGiavwv
3x the credits, same $29
The Creator plan now includes:
• 600 credits a month, with rollover
• 30-minute videos (vs 1 min on Free)
• 1080p exports, no watermark
Sign up for the new Creator plan today + reply "creator," and we'll drop 100 bonus credits in your account
Build Something @Lovable With @HeyGen@HeyGen is now built directly into @Lovable, and most of my feed is treating it as a minor update. It is not.
What actually shipped is this:
You can now build an app in Lovable and create the talking AI video for it inside the same tool.
The avatars, the voices and the video all live right in the build, with no second product to learn and no exporting and re-importing between them.
This is not a one-off either.
HeyGen already shows up in or alongside ChatGPT, Claude, Canva, Gamma, HubSpot and Linear, and now Lovable joins the list.
The pattern is obvious once you see it, because HeyGen is making itself the video button inside almost every tool you already work in.
Here is the part worth slowing down for.
For years the hard part of video was the production itself, the camera and the studio and the edit that ate your weekend.
That has collapsed into a single click, which means the real question has moved.
It is no longer whether you can make the video, it is whether you have anything worth saying.
That is great news if you do, and it is a problem if "I do not have the setup" was quietly your reason for not shipping, because that reason is gone now.
The tools just took away the last excuse you had left.
@HeyGen this is the smartest distribution play in AI video right now.
You are already in or alongside ChatGPT, Claude, Canva, Gamma, HubSpot and Linear, and now Lovable. What is the next tool on the list?
@HeyGen You're only seeing the close-up. 🧊
@HeyGen ...
you wanted these for the Christmas party, so I iced the whole thing out.
Full video dropping soon.
Your SWAG is officially in production. 😎😏
GET 58% OFF EMERGENT :
Emergent is ONE year old.
10 million builders. 58% off. $20 → $9/month.
Annual plan. Locked in.
Alongside Bolt and Rork...
one of the best vibe coding apps out right now.
And the most aggressively priced.
Mobile app dev.
Private hosting.
GitHub.
Credits that scale.
10 million builders is signal.
People are already shipping real work on this platform.
At $9 you're not buying a subscription.
You're buying a year of reps.
Don't overthink it. 👇
https://t.co/kGG0UGL2qW
Avatar V now takes direction
Custom Motion let you cue specific motions, expressions and gazes
→ A lean-in before a key point
→ A smile for emphasis
→ A thumbs up at the end
Whatever the moment calls for, just type it