New weekly update workflow Step 1: open Claude Cowork Step 2: "read this week's Slack and make me a video using the HeyGen MCP" Step 3: pick from 3 scripts Step 4: watch your avatar deliver it Stop digging through Slack for what you missed. Watch the update instead

Get your AI video right, down to the word. Full control. You can now read the full composition of a Video Agent video, scene by scene, and edit it granularly. One scene changes, the rest stay as they were. Build it into a repeatable pipeline with the edit scenes endpoint, and scale from there. developers.heygen.com/docs/e…

if you tried hyperframes, you would know the agentic video stack is being built on code-gen but SWE-bench code v.s. code-to-video that feels alive are two different things we built the Code2Video Bench, in collab w/ Google DeepMind & @Kaggle frontier labs can finally get good at agentic video tasks

muse is great!! made this video using Hyperframes + HeyGen MCP to showcase Muse and its capabilities, @alexandr_wang @Muse

Had some free time, so I made a HeyGen MCP video using Hyperframes + HeyGen MCP.

Medicine is the only job where the expert with the most to say has the least time to record it. HeyGen runs inside ChatGPT and Claude, so you can turn the notes you already wrote into a month of clinic videos, without a camera or a free afternoon.

excited for Raycast v2 - it deserve more 👀 so we built a @HyperFrames_ video agent for you, it can cook on-brand launch vids for you 24/7 u can finally launch as fast as you ship code claim it here: hyperframes.dev/company/rayc…

Excited to be a launch partner for GPT-Live-1 with @OpenAI. Building this demo together was a blast. We also open sourced the whole demo, see link below. Why it matters: GPT-Live-1 is one model that listens and speaks at the same time. No STT to LLM to TTS chain, no brittle handoffs. It handles interruptions, background noise and mid-sentence changes of mind, and hands off deeper reasoning to a backend model while the conversation keeps flowing. Pair that with LiveAvatar and the voice gets a face that reacts in real time. Full duplex audio plus a live avatar is as close to talking to a person as AI gets today. openai.com/index/introducing…

We open sourced the demos from the video so nobody has to start from an empty repo. If your idea needs a realtime experience, this is the place to start. GPT-Live-1 for speech, LiveAvatar for the face, tool calls landing on screen as overlays. Get started: github.com/heygen-com/liveav…

HeyGen Professional Voice Clone Voice has been the biggest gap in lifelike avatars, until now. A dedicated voice model, trained on twenty minutes of one speaker, gives your avatar a voice that truly sounds like you. Now available via API: developers.heygen.com/docs/v…

Update from HeyGen research team: in Video Podcast, every turn is generated on its own. Our new technical report details how continuity is enforced across them - frame-aligned durations on a 25 fps grid, camera views generated from one shared scene, and a mouth-movement check on every listener cutaway. heygen.com/research/video-po…

API pricing update worth your attention: a managed real-time avatar now costs less per active minute than the raw GPU math of self-hosting one. Come build with us: liveavatar.com

New from HeyGen and the Google Cloud team: how we brought our diffusion model for avatars, Avatar IV, to Trillium (v6e) TPUs. Six optimization milestones, start to finish: heygen.com/research/heygen-x…

Introducing the HyperFrames MCP in Claude Design The difference between prototyping a design and shipping it as video Animate in Claude Design, hit share, and hyperframes agent scores it, adds sfx, tightens the motion and renders the MP4 Just send it to HyperFrames. We'll take it from there.

New from HeyGen Engineering: how we replaced our custom job orchestrator with Temporal to power every video we render. Millions of workflows a day. Tens of millions of activities. 300+ workflow types across 50+ worker fleets on multiple clouds. Read more: heygen.com/research/how-temp…

Ship the take. Not the ums! Filler Word Removal API is live: one call takes a raw recording and returns it clean - fillers cut, dead air gone, no manual editing. API docs → developers.heygen.com/filler…

Wow @HyperFrames_

A story that made my week. A month ago, a dad wrote to our support team. His 5th grader's robotics team was building a talking hologram of a historical figure for the World Robot Olympiad. They needed @TryLiveAvatar to bring it to life. 39 emails over a month: billing questions, consent for a historical figure, a last-minute avatar redo because hologram projection needs a pure black background. The kind of thread that could easily fall through the cracks. It didn't. A human support specialist and an internal @openclaw agent ran it as one team. When one was offline, the other picked up. No question ever waited until morning. This week he wrote back: 2nd out of 400+ teams at regionals, advancing to nationals. Plus a handwritten thank-you note from the kids. Our favorite builders this week are in 5th grade.

The take was perfect. The audio wasn't. Swap the audio, keep the take. Lipsync API re-syncs the lips to match. One call. Precision mode. Lipsync API → developers.heygen.com/lipsyn… Batch → developers.heygen.com/batch-…