Synthesia launches Express-3 video model, halving avatar video creation time and adding cartoon Style Avatars
alexvoica · x · 2026-09-10
Synthesia has unveiled Express-3, a new video model that produces sharper, more expressive AI avatars while cutting video creation time in half.
- Style Avatars: Beyond classic photorealistic avatars, Express-3 supports a new category of stylized, cartoonish characters generated from a prompt — a long-requested feature aimed at company mascots and more accessible content.
- Synthesia Assistant integration: Generate an avatar, include it in a video, and the Assistant automatically matches the visuals to the avatar's characteristics.
Related event: Synthesia Launches Express-3 Avatar Model(2 posts)→
More from Multimodal
- Adobe ships new Photoshop AI features: Light Adjustment Layer, Instruct Edit with Masks powered by Firefly Image 5 — rufusd · 2026-09-10
- Ex-OpenAI Sora Head Bill Peebles Teams With Katzenberg on Hollywood AI Startup — steph_palazzolo · 2026-09-10
- VRAM guard trick: run a local LLM and ComfyUI video generation side by side without OOM — Sn0opY_GER · 2026-09-10
- Trained a 210M text-to-image diffusion transformer from scratch on one GPU in 3.5 days — IvanMikhnenkov · 2026-09-10
- HF rebuilds AUTOMATIC1111 as a 73-node Gradio Workflow with 11 pipelines, no 20GB downloads — Gradio · 2026-09-10
- An interactive journey through the 3D ecosystem: why now is the best time to start — SergioPaniego · 2026-09-10