pickfit

BrowseDeveloper CLIs

Kesha Voice Kit

Local speech-to-text and text-to-speech toolkit for agents and scripts.

Tested on Linux

Use it

Linux x86_64

You need first

  • Runtimes: Bun>=1.4.0
  1. Download Bun installer script only

    curl -fsSL https://bun.sh/install -o /tmp/bun-install.sh && ls -l /tmp/bun-install.sh && head -c 500 /tmp/bun-install.sh
  2. Install Bun 1.4.4 via downloaded script

    bash /tmp/bun-install.sh && export PATH="$HOME/.bun/bin:$PATH" && bun --version
  3. Install Kesha via bun global add

    export PATH="$HOME/.bun/bin:$HOME/.local/bin:$PATH" && bun add -g @drakulavich/kesha-voice-kit

Check that it works

export PATH="$HOME/.bun/bin:$HOME/.local/bin:$PATH" && kesha --version

The output should contain “2.4.0”.

First thing to try

export PATH="$HOME/.bun/bin:$HOME/.local/bin:$PATH" && kesha status

Run kesha status to check the install and see what is set up.

What it does

  • Connect voice tools to agents over MCP
    How to use it
    1. Start the MCP server with kesha mcp, which reads JSON-RPC messages on standard input.
    2. Point your MCP client at kesha mcp, or pipe requests in with kesha mcp < <file>.
    3. Send tools/list to see the available tools, such as synthesize_speech.
    Tested
  • Transcribe audio to text on your machine
    Not tested
  • Turn text into speech
    Not tested

How far pickfit got

Linux x86_64

  1. Installs
  2. Starts
  3. 1 of 3 features work
  4. Works offline

macOS and Windows: the project says it runs there; pickfit hasn’t tried it.

Good for

  • transcribe an audio file to text
  • synthesize speech from text to a wav file
  • expose transcription and synthesis to MCP clients
  • dictate from microphone to text

An alternative to Whisper, Whisper large-v3-turbo and faster-whisper.

How pickfit tested it

Each run starts from a clean machine and follows the steps above. How pickfit verifies

Linux x86_64

Its main function works

  • Replayed from scratch
  • Works offline
  • Checked on 30 Oct 2026
  • Version 2.4.0

Differences the test machine hit

  • README: Runtime: Bun >= 1.4.0
    In the test: install Bun 1.4.4 via bash /tmp/bun-install.sh and export PATH="$HOME/.bun/bin:$PATH" for non-interactive shell
  • README: curl -fsSL https://bun.sh/install | bash
    In the test: installer fetches bun-linux-x64.zip binary from github.com/oven-sh/bun which cannot be read as text before running
  • README: kesha install pulls ~2.5GB on Linux/Windows and is always explicit
    In the test: kesha install needs 2.7GB cold-cache download (67MB engine + 2.5GB ASR + 86MB lang-id) before transcribe/synthesize work; kesha status/install --plan work without it

How to read a tool card

Stitching
pickfit installed this tool itself. Everything under the stitches comes from that run. A tool pickfit has not run never gets stitches.
Verified on your OS
The install steps ran on the system you chose. “Ran on Linux only” means it has not been run on yours yet.
macOS, Linux and WindowsWhere it ran
Green OS marks: pickfit ran the install steps there. Grey ones: the README says it runs there, and pickfit has not tried.
How far the run got
Installs and starts, its main function works, its service answers, it produces model output. A window means a desktop app opened; a plug, an MCP client used it. The mark sits beside the run’s result, under “Why it fits and how we know” and in “How pickfit tested it”.
Visual check
pickfit looked at screenshots of the app. Green: it looks right. Amber with a number: things that look wrong; open it to see them. Grey and dashed: not checked yet.
Known vulnerabilities
Known high or critical vulnerabilities (OSV.dev) in what it installs. Green: none known. Red with a number: how many; open it for the list, and read it before you install. Grey and dashed: not checked, which does not mean safe.
About the run
The day it ran, and the version that ran.
Source of a reason
Every reason a pick fits rests on one profile field. Open the reason to see the field and the value it relies on.
Hidden gem
High quality, little known.
Quality and popularity
Two scores out of 100. Quality: how well it installed and worked when pickfit tried it, and how good its install docs and upkeep are. Popularity: from stars, downloads and Hacker News.
Features checked
Of the main features the README advertises, how many the run actually exercised (green), and how many failed (red). Grey ones were not checked, and the card says why.
Not verified
A well-known tool Fitz mentions from general knowledge. No run record, no steps, no score.