Chatterbox TTS Server
devnen/Chatterbox-TTS-ServerReviewed by AI
Checked by RepoPicks Radar on
What is Chatterbox TTS Server?
This project wraps Resemble AI's Chatterbox speech models in a server with a web page and an API. You paste text, choose one of the included voices or upload a short recording to copy a voice, and it writes out audio; long texts are split automatically, so a whole book becomes one file. Three models are offered (English, 23 languages, and a faster Turbo one) and you switch between them in a dropdown. It runs on Windows, Linux and Apple Silicon with Python 3.10, faster with an NVIDIA or AMD graphics card, and there is also a Docker route and a Google Colab demo.
Chatterbox TTS Server is free and open source (MIT) and can replace ElevenLabs, Murf, and Speechify.
How easy is it to install Chatterbox TTS Server?
C · install 1 or 2 tools · Windows
- Windows C
- macOS D
- Linux C
Windows is the easiest: install Python 3.10, clone the repository and double-click start.bat, and the launcher detects your card, installs everything and opens the web page (its Portable Mode then makes the folder movable). Linux is the same work with ./start.sh. macOS is harder (a D): the hardware table lists Apple Silicon as "Manual install", so you build the environment and install the packages by hand.
Chatterbox TTS Server on GitHub
- Stars
- 1,499on Nov 10, 2026
- Downloads
- —all releases
- Growth
- +4stars in 6 days
Stars over time
Facts
- Runs on
- Windows, Linux, macOS
- You need to install
- Python, Docker, NVIDIA graphics card, FFmpeg
- To build from source
- Python, Docker, NVIDIA graphics card, FFmpeg
- Disk space
- 950 MB–4.3 GB · estimate
- Accounts or keys
- Hugging Face
- Cost
- Free · The server and the models are free to download and run on your own machine; you pay in hardware and electricity, or in rented GPU time.
- License
- MIT (Permissive)
- Can replace
- ElevenLabs, Murf, Speechify
- Last change on GitHub
- May 26, 2026
Chatterbox TTS Server risk signals (6)
- AttentionDependencies with known vulnerabilities
Pinned dependencies with known vulnerabilities (OSV): torch: GHSA-63cw-57p8-fm3p, GHSA-qfhq-4f3w-5fph, GHSA-rrmf-rvhw-rf47; diffusers: GHSA-7wx4-6vff-v64p, GHSA-98h9-4798-4q5v, PYSEC-2026-2446; transformers: GHSA-29pf-2h5f-8g72, GHSA-37mw-44qp-f5jm, GHSA-4w7r-h757-3r74; torch: GHSA-rrmf-rvhw-rf47, PYSEC-2026-139; diffusers: GHSA-7wx4-6vff-v64p, GHSA-98h9-4798-4q5v, PYSEC-2026-2446; transformers: GHSA-29pf-2h5f-8g72, GHSA-37mw-44qp-f5jm, GHSA-4w7r-h757-3r74; torch: GHSA-3749-ghw9-m3mg, GHSA-53q9-r3pm-6pq6, GHSA-63cw-57p8-fm3p; diffusers: GHSA-7wx4-6vff-v64p, GHSA-98h9-4798-4q5v, PYSEC-2026-2446; transformers: GHSA-29pf-2h5f-8g72, GHSA-37mw-44qp-f5jm, GHSA-4w7r-h757-3r74; diffusers: GHSA-7wx4-6vff-v64p, GHSA-98h9-4798-4q5v, PYSEC-2026-2446; transformers: GHSA-29pf-2h5f-8g72, GHSA-37mw-44qp-f5jm, GHSA-4w7r-h757-3r74; torch: GHSA-3749-ghw9-m3mg, GHSA-53q9-r3pm-6pq6, GHSA-63cw-57p8-fm3p; diffusers: GHSA-7wx4-6vff-v64p, GHSA-98h9-4798-4q5v, PYSEC-2026-2446; transformers: GHSA-29pf-2h5f-8g72, GHSA-37mw-44qp-f5jm, GHSA-4w7r-h757-3r74; torch: GHSA-3749-ghw9-m3mg, GHSA-53q9-r3pm-6pq6, GHSA-63cw-57p8-fm3p; diffusers: GHSA-7wx4-6vff-v64p, GHSA-98h9-4798-4q5v, PYSEC-2026-2446; transformers: GHSA-29pf-2h5f-8g72, GHSA-37mw-44qp-f5jm, GHSA-4w7r-h757-3r74
automatic check
- Attentionengine installed from a git fork
Every install path ends with "pip install --no-deps git+https://github.com/devnen/chatterbox-v2.git@master" (Dockerfile and all requirements files). The speech engine is pulled from the maintainer's own fork at whatever the master branch holds that day, and --no-deps skips the dependency checks, so what lands on your machine is not pinned to a version.
AI review
- Attentionvoice cloning of real people
"Voice Cloning: Supports voice cloning using a reference audio file (.wav or .mp3)" with no consent or acceptable-use text anywhere in the README. A short recording of someone's voice is enough to make them appear to say anything. The engine does include Resemble's watermarking library (resemble-perth==1.0.1 in the requirements files).
AI review
- Attentionweb server without a password
The server opens port 8004 with a web UI and an OpenAI-compatible API, and the UI can "view/edit config.yaml settings (server, model, paths)"; the README describes no login or token. The v2.0.0 notes also record a past hole: "Security: CWE-22 path traversal fixed on /tts and /v1/audio/speech voice file parameters". Keep it on your own machine or behind something that asks for a password.
AI review
- Infopython 3 10 only
README prerequisites: "Python: Version 3.10 required ... Python 3.11+ is not supported — key dependencies lack pre-built wheels, causing build failures." On Windows the Portable Mode downloads its own embedded Python 3.10.
AI review
- Infoamd docker relaxes container isolation
docker-compose-rocm.yml, -rdna4.yml and -strixhalo.yml set "security_opt: seccomp=unconfined", "ipc: host" and device access to /dev/kfd and /dev/dri, and suggest uncommenting "privileged: true" and "/dev/mem" if GPU access fails, which gives the container far-reaching access to the host.
AI review
Evidence is shown in its original language.
What Radar could not check about Chatterbox TTS Server
- Whether it installs and uses your particular graphics card, given the many separate requirement files per card family
- Real disk use: the README itself recommends 10 GB or more for dependencies and model cache, and the models alone are about 2 GB
- How good the cloned voices sound and how fast generation is on your hardware
- Whether every generated file carries the Perth watermark
- Whether the Google Colab demo still runs