Self-host Speaches with Levelrail

An OpenAI-compatible speech server: Whisper transcription and text-to-speech behind one API.

  • Authorspeaches-ai
  • LicenceMIT
  • Memory2 GiB
  • Archarm64
  • Services1

How to self-host Speaches

  1. Pull the images

    One image is pulled on your server.

    Show images
    • ghcr.io/speaches-ai/speaches:0.8.3-cpu
  2. Create the app

    One service starts in a single Compose app, with 1 named volume mount for your data, on port 8000.

  3. Fill in the secrets

    1 setting is preset, you do not have to fill in anything.

  4. Open it

    Without a domain the app gets an sslip.io address; add --domain SERVICE=HOST to use your own.

Speaches requirements

Memory
About 2 GiB
Architectures
amd64 image published, arm64 image published
Ports
8000
Domain
Optional, an sslip.io address is used without one
Settings
none to fill in
License
MIT, permissive license
  • Listed. It is in the catalog but not verified. Back up the data volume before changing an image tag.
  • Host ports. It publishes port 8000 on the host, so open them in the VPS firewall only if you want that.

Install Levelrail first

On a server with Levelrail running, run the command above, or open the dashboard: Apps, New app, Browse templates, then Speaches. New to Levelrail? Install it on a fresh VPS first.

docker pull ghcr.io/glincker/levelrail:latest

Then levelrail-cli login --device. See Getting started and the Speaches docs page.

Deploy Speaches your way

Every format below deploys the same pinned images. Pick the one that fits how you work.

Deploy Speaches from the CLI

levelrail-cli templates deploy speaches --name my-speaches

Deploy Speaches with the API

Create a token in the dashboard first. This one-click route deploys it as a trial app you can keep or remove.

curl -X POST -H "Authorization: Bearer $TOKEN" \
  https://levelrail.example.com/api/v1/service-templates/speaches/deploy

Speaches configuration

Speaches services

ServiceImagePortsVolumes
speaches ghcr.io/speaches-ai/speaches:0.8.3-cpu 8000 speaches_models -> /home/ubuntu/.cache/huggingface/hub

Speaches environment variables

VariableServiceValue
ENABLE_UI speaches Preset by the template

Ports and volumes

Ports
8000
Volumes
1 named mount: speaches_models -> /home/ubuntu/.cache/huggingface/hub

Speaches FAQ

How much memory does Speaches need?

The catalog lists about 2 GiB of memory for Speaches.

Does Speaches run on arm64?

For Speaches, an arm64 image is published; amd64 is image published.

What licence is Speaches under?

Speaches is licensed under MIT, which is permissive, with few conditions on reuse.

Does Speaches need configuration before it starts?

No. The template fills in its own defaults.

Which ports does Speaches use?

Speaches listens on port 8000 inside the stack.

Where does Speaches keep its data?

In a named volume: speaches_models mounted at /home/ubuntu/.cache/huggingface/hub.

Can I deploy Speaches on Dokploy or Dokku?

The services and pinned images on this page can be run on any Docker host. The public Dokploy templates list we compared does not have an entry with the same id for Speaches, so on Dokploy you would use a Docker Compose application with the file from this page. Dokku deploys container images with its own commands, so the pinned image ghcr.io/speaches-ai/speaches:0.8.3-cpu is what you deploy there.

levelrail-cli templates deploy speaches --name my-speaches