persona-2.0-live just doubled its resolution.
2× the resolution, 4× the pixels. The same word-level control over expression.
Try it now: keyframelabs.com
In preview now for all customers, through the platform and the API.
Agents are now first-class in the Keyframe API. Agent IDs, authentication, dynamic variables and more.
keyframelabs.com/changelog/a…
An upgraded usage analytics experience is now live on the Keyframe platform. Monitor sessions, minutes, concurrency and more for every agent and avatar. keyframelabs.com/changelog/u…
Teams can now share a workspace in Keyframe. Invite your team to an organization, set each member's role, and collaborate on agents, avatars, API keys and more.
keyframelabs.com/changelog/o…
Our favorite setup right now: @ElevenLabs v4 + persona-2.0-live. The best of both worlds!
Just add [emotion] tags to your existing voice or agent prompts, and persona-2.0-live handles the rest. For finer-grained control, you can always use our SDK.
Introducing the world's most expressive and controllable avatar model.
Precise, word-level control over expression, combined with industry-leading realism and scalability.
persona-2.0-live is now in preview. Try it now: keyframelabs.com
Repo: github.com/keyframelabs/exam…
Full written guide: keyframelabs.com/blog/system…
Sign up: platform.keyframelabs.com
Docs: docs.keyframelabs.com
🔨 & ☕: The best interviews feel more like collaboration than evaluation.
Watch @TheTmartz take on TinyURL with Lyra, an AI interviewer that works through the architecture with him in real time over a shared canvas.
Avatar by Keyframe. Voice by @ElevenLabs.
Repo, guide and more ↓
Replying to @dflieb
Built in a day with persona-1.5-live. New backchanneling feature means the avatars react while the other is speaking (first of its kind). More to come!
Join our new discord community and share your thoughts! discord.gg/Wfce5xHNd5
