kokoro v0.1.0

Kokoro-FastAPI — a fast, OpenAI-compatible text-to-speech API server powered by the Kokoro model, with multiple voices and a small web UI. Models are bundled in the image. Deployed as a single stateless host-networked Nomad service.

nomad-pack run kokoro --registry nomploy

…or one line (add registry + run)

curl -fsSL https://packs.nomploy.com/install.sh | sh -s -- kokoro

Needs nomad-pack on PATH. The script only adds the nomploy registry and runs this pack.

Source ↗ Project ↗ ★ 5.5k ⚑ Report an issue

1 task http 8880 image ghcr.io/remsky/kokoro-fastapi-cpu:latest tracks :latest image bumped today

Variables 7

values.hcl

Save as values.hcl, edit, then run:

nomad-pack run kokoro -f values.hcl --registry nomploy

# The name of the Nomad job.
job_name = "kokoro"

# The Nomad namespace to deploy into.
namespace = "default"

# The datacenters to deploy to.
datacenters = ["*"]

# The Kokoro TTS container image. Pin a tag in production.
image = "ghcr.io/remsky/kokoro-fastapi-cpu:latest"

# Host port for the Kokoro TTS API / web UI. Fixed at 8880 inside the image.
port = 8880

# Placement constraints. On a nomploy cluster: attribute = "$${meta.nomploy_control_plane}", operator = "=", value = "true".
constraints = []

# The task resources.
resources = {
    cpu    = 2000
    memory = 2048
  }

Name	Type	Default	Description
job_name	string	"kokoro"	The name of the Nomad job.
namespace	string	"default"	The Nomad namespace to deploy into.
datacenters	list	["*"]	The datacenters to deploy to.
image	string	"ghcr.io/remsky/kokoro-fastapi-cpu:latest"	The Kokoro TTS container image. Pin a tag in production.
port	number	8880	Host port for the Kokoro TTS API / web UI. Fixed at 8880 inside the image.
constraints	list	[]	Placement constraints. On a nomploy cluster: attribute = "$${meta.nomploy_control_plane}", operator = "=", value = "true".
resources	object	{ cpu = 2000 memory = 2048 }	The task resources.

Readme

kokoro

Kokoro-FastAPI — a fast, OpenAI-compatible text-to-speech API server powered by the open Kokoro model. Point any client that speaks the OpenAI /v1/audio/speech API at it to generate natural speech in multiple voices — great for self-hosted assistants, audiobooks and accessibility.

Single stateless host-networked Nomad service (models are bundled in the image; no volume needed). This pack uses the CPU image.

Deploy

nomad-pack registry add nomploy https://github.com/Nomploy/nomad-packs
nomad-pack run kokoro --registry=nomploy

Configure

Variable	Default	Description
`port`	`8880`	API / web UI port. Fixed at `8880` inside the image.
`image`	`ghcr.io/remsky/kokoro-fastapi-cpu:latest`	Container image. Pin a tag in production.
`resources`	`{ cpu = 2000, memory = 2048 }`	Task resources. CPU inference is heavy — give it cores.

Endpoints: API at /v1/audio/speech, interactive docs at /docs, a demo UI at /web. Works as a drop-in TTS backend for open-webui and other OpenAI-compatible clients. For much faster synthesis on an NVIDIA GPU, switch image to the kokoro-fastapi-gpu tag and add GPU scheduling.