Pipecat Simli max_session_length vs Sume's 4 to 60 second clip
Simli in Pipecat caps a live session with max_session_length and max_idle_time. Sume has no session: an avatar job is a 4 to 60 second file.
What limits does Pipecat's Simli service set?
Pipecat describes Simli as real-time avatar video over WebRTC, animated by your agent's audio. The service takes an api_key and a face_id. It exposes max_session_length, the absolute maximum session duration in seconds, and max_idle_time, the maximum idle time before disconnect. Audio is resampled to 16 kHz internally.
What is the Sume equivalent?
There is no session on Sume, so there is nothing to disconnect. The limit is on the job: the target video duration, estimated from your script, must be 4 to 60 seconds inclusive. Longer scripts are rejected; split them into more jobs.
The aspect ratio is 1:1, 3:4, 9:16, 4:3 or 16:9 (default 9:16), at 720p.
| Limit | Pipecat Simli | Sume avatar job |
|---|---|---|
| Maximum length | max_session_length, set by you | 60 seconds per job |
| Minimum length | Not stated | 4 seconds |
| Idle behavior | max_idle_time then disconnect | None; the job ends when the clip is done |
| Frame format | WebRTC video | MP4 at 720p |
How do I stretch past 60 seconds?
Split the script at sentence breaks into scenes of under 60 seconds and join the finished clips with Timeline. A Timeline render takes up to 200 video slots and bills $0.10 per output minute, so a three clip, 2 minute video is $0.30 to join, plus the avatar jobs.
For a pause where the avatar waits, use a silence voice input with a duration, which is covered in the silence beats post.
What should I check before choosing?
- Does the viewer need to interrupt? Then a live service fits and Sume does not.
- Is the script fixed? A recorded clip is cheaper per view and reviewable.
- Is the length under 60 seconds per piece? Then one job per piece works.
Sources
Related posts
More in Sume Avatar 1.0
- Pipecat TavusVideoService live avatar, or a recorded Sume clip?
Pipecat's Tavus service speaks your agent's TTS live over WebRTC. A Sume avatar job is a 4 to 60 second recorded MP4. How to pick, by what the viewer does.
- Reuse one AI presenter across lip-sync clips with an avatar handle
Create the presenter once, then pass avatar_handle to H3 Max lip sync for every clip. Same face each time, no re-upload, one Idempotency-Key per line.
- Tavus Griffin-Lite is invite-only: what can you build today?
Griffin-Lite is a closed Tavus preview. Until you get in, Sume ships rendered avatar clips, lip sync and motion control for talking-head video.
- YouTube Shorts series: a weekly AI presenter episode pipeline
YouTube began rolling out Shorts series on 2026-09-23. A repeatable weekly pipeline for presenter episodes: one avatar, draft at standard, final at max.
Written by Sume