0.2.16

QuentinFuxa/WhisperLiveKit0.2.16Nov 25, 2025by QuentinFuxa

AI Summary

Significant scalability improvements are achieved by sharing the Whisper decoder across multiple sessions, removing a previous bottleneck.

Key Highlights

A single Whisper decoder can now be shared across sessions/users, removing scalability bottlenecks.
Improved performance by sharing the encoder (which was already shared).

Breaking Changes

Removed the `--preload-model-count` parameter as it is no longer needed.

New Features

Shared Whisper decoder across sessions
SimulStreaming/AlignAtt policy improvements

Full Release Notes

## Core
- SimulStreaming / AlignAtt policy: a single Whisper decoder can now be shared across sessions/users (the encoder was already shared). This removes several scalability bottlenecks. Co authored with [@eschmidbauer](https://github.com/eschmidbauer). Therefore, `--preload-model-count`  parameter is now useless and has been removed.

## Bug fixes
- Correct silence handling when language is set to `auto` but not yet detected
- Fix silence detected but never reported by Silero VAD
- Add troubleshooting docs for #271, #276, #284, #286