whisper_streaming and SimulStreaming (UFAL)
Research implementations that turn Whisper into a streaming transcriber. whisper_streaming (LocalAgreement policy, 2023) is now marked outdated by its authors in favour of SimulStreaming.
Overview
Best for: Researchers and builders who want the underlying streaming policies to embed in their own server.
At a glance
Research code over raw TCP/stdin; papers report a few seconds of latency. Languages follow Whisper (about 99). whisper_streaming is superseded by SimulStreaming.
16 kHz mono audio via the provided server/clients.
n/a
Whisper languages.
Configurable chunk/min-chunk; papers report a few seconds of latency on GPU (not re-verified).
Self-hosted.
Self-hosted.
GPU recommended for large Whisper models.
MIT (both repos per GitHub metadata).
Features
- streaming transcription and translation
- LocalAgreement or simultaneous decoding policies
- VAD option
Pricing
| What | Price | Unit |
|---|---|---|
| Software | $0 |
Self-host compute only.
Free tier: Open source.
Source: github.com
Setup
- git clone https://github.com/ufal/SimulStreaming
- Install requirements and follow README to run the server and stream audio.
- Prefer wrappers such as WhisperLiveKit for a ready server.
Endpoint
Local TCP server (see README)Authentication
none
Warnings
Original repo is outdated
The whisper_streaming authors say it is being replaced by SimulStreaming; start new work on SimulStreaming or a wrapper that bundles it.
Not production packaging
These are research implementations without auth, multi-tenant scaling or monitoring.
Small community for SimulStreaming
SimulStreaming has far fewer stars and contributors than WhisperLive or WhisperLiveKit.
Plus 3 warnings that apply to all open models APIs. See category warnings.
Limits
- Research-grade code: one stream per process, minimal ops tooling.
Models and products
| Name | Status |
|---|---|
| whisper_streaming | Superseded |
| SimulStreaming | Active |
Docs and sources
Docs
Sources used
- raw.githubusercontent.com/ufal/whisper_streaming/main/README.md
- api.github.com/repos/ufal/whisper_streaming
- api.github.com/repos/ufal/SimulStreaming
Exact run commands and latency figures (no snippet given).