Kokoro-FastAPI---CPU

Dockerized FastAPI wrapper for Kokoro-82M text-to-speech model w/CPU ONNX and NVIDIA GPU PyTorch support, handling, and auto-stitching.

  • AI
  • Other
  • Productivity
  • Tools & Utilities
Dockerized FastAPI wrapper for Kokoro-82M text-to-speech model w/CPU ONNX and NVIDIA GPU PyTorch support, handling, and auto-stitching.
This is a version meant for running on CPUs. It is not recommended to run this on a CPU unless you have a very powerful CPU, as it will be slow. If you have a GPU, please use the GPU version of this container.

Platform availability

Available on 2 of 10 platforms

  • Unraid
  • TrueNAS (not listed)
  • Umbrel
  • ZimaOS (not listed)
  • Proxmox (not listed)
  • Helm (not listed)
  • Mac (not listed)
  • Windows (not listed)
  • Linux (not listed)
  • Any server (not listed)

Health score?A 0-100 score computed nightly from four factors: maintained (40%: recent commits and steady activity), popular (25%: GitHub stars, log scale), easy to install (20%: platforms, install notes, setup guides, Docker image) and light to run (15%: minimum RAM, ARM support). When the source code has been checked, a fifth factor, security check (20%: automated scan for committed secrets, unsafe desktop-app settings, known-vulnerable dependencies and project hygiene; any serious finding caps it at 20), is added. Factors without data are left out and the rest are rescaled.

72/100 Good

Maintained
100last commit 17 days ago
Popular
755,633 GitHub stars
Easy to install
452 install platforms, install notes, Docker image
Light to run
304 GB minimum RAM

Availability: platform listings confirmed today, links checked 21 days ago

GitHub stars
5,633
Open issues
15
Last commit
2026-11-04
Latest release
v0.9.2
License
Apache-2.0
Activity
rising

Checked 17 days ago - source: GitHub (remsky/Kokoro-FastAPI)

Resources & Compatibility

Minimum requirements: 4 GB RAM, 2 CPU core(s).

Checked 54 days ago - source: Docker Hub image tags and published docs

First-install notes

  • Install the Kokoro JS package Run `npm i kokoro-js` to add the Kokoro JavaScript library to your project. If this is missing, the app cannot load the TTS model or make audio. source
  • Pick a model and compute device Set model_id to a valid Hugging Face model ID and choose the compute device (where the app does the math), such as cpu for Node or wasm in a browser. Picking the wrong device or an unavailable model ID can stop Kokoro from starting. source
  • Choose a voice for every speech Pass a voice identifier, like af_heart, when generating audio, and call save with a .wav file name if you want the sound written to disk. Missing a valid voice or forgetting to save can leave you with no usable speech file. source

Fetched 53 days ago - each note links to its source

Alternatives

No alternatives collected yet.

Common questions

Will the CPU-only build be fast enough for normal use?

It is not recommended for CPUs unless the CPU is very powerful, because it can be slow. If you have a GPU, use the GPU version instead.

Why can another container on the same Docker network not reach the service at localhost?

Try the host machine’s IP address from within the Docker network. If that still fails, check the expected port, whether the containers can see the host, and whether traffic filtering is blocking it.

What should I do if the Unraid app template fails?

If the Unraid app template does not work, a user found success by starting it with Docker run or Docker Compose instead.

Answers sourced from forums.unraid.net, github.com, github.com

Community

Collected 49 days ago - public community threads