CHECKED OCT 10, 2026
GA Open source (QuentinFuxa)

WhisperLiveKit

Popular self-hosted realtime transcription server with a web UI, combining SimulStreaming / WhisperStreaming policies with optional streaming speaker diarization (Sortformer).

Est. per minute$0

Overview

Best for: Quick self-hosted live transcription with a usable web UI and optional diarization.

At a glance

TypeSpeech-to-text
LicenceApache-2.0
CommercialYes
StreamingYes
Languages99

Whisper models with SimulStreaming or LocalAgreement policies; languages follow Whisper (about 99). Optional streaming diarization.

Audio in

Browser mic via built-in web page or WebSocket clients.

Audio out

n/a

Languages

Whisper languages.

Latency

Policy and model dependent.

Regions

Self-hosted.

Compliance

Self-hosted.

Hardware

GPU recommended beyond the base model; Apple Silicon supported via MLX backends (per project, not verified).

Licence

Apache-2.0

Features

  • streaming partials
  • realtime speaker diarization (Streaming Sortformer option)
  • web UI
  • model management CLI (wlk pull/rm/models)
  • file transcription to SRT

Pricing

WhatPriceUnitNotes
Software$0Compute only.
How the per-minute estimate was worked out

Self-host compute only.

Free tier: Open source.

Source: github.com

Setup

  1. pip install whisperlivekit
  2. wlk --model base --language en
  3. Open the local web UI and allow microphone access.

Endpoint

http://localhost:8000 (default UI; check CLI output)

Authentication

None by default.

Quick start bash

pip install whisperlivekit
wlk --model base --language en      # starts server + web UI
wlk transcribe --format srt podcast.mp3 -o podcast.srt   # offline file mode

Written from the current docs. Check the vendor's SDK version before you ship.

Warnings

No built-in auth

Run it behind a reverse proxy with authentication before exposing it beyond localhost.

Diarization adds GPU load

Streaming diarization models run alongside Whisper; size the GPU accordingly.

Default port unverified

Check the CLI output for the actual port before wiring clients.

Plus 3 warnings that apply to all open models APIs. See category warnings.

Limits

  • Single-server design; scale-out is up to you.

Models and products

NameStatusNotes
Whisper models (tiny..large-v3) with SimulStreaming or LocalAgreement policiesActive projectLast push 2026-10-01; CLI 'wlk'.

Docs and sources

Docs

Sources used

Not fully verified

Default port, MLX support.

Similar open models APIs

NVIDIANVIDIA Nemotron ASR Streaming / Parakeet (Riva, NIM, NeMo)$0/min0 high-severity warningsKyutaiKyutai STT (Delayed Streams Modeling)$0/min0 high-severity warningsMoonshine AI (Useful Sensors)Moonshine Voice (Moonshine v2 streaming)$0/min0 high-severity warningsCollabora (open source)WhisperLive (Collabora)$0/min0 high-severity warnings
Spotted a wrong price or a dead link?