Piper CLI
Piper is a CLI for local text-to-speech, with downloadable voices, WAV output and options for your own speech applications.
- 6,183 ★ on the repository
- Last commit
- GPL-3.0
Extended access
Why this level
- The CLI downloads models and creates audio files; playback mode emits sound on the device.
Install
In your terminal, with SkillFoxx CLI
npx skillfoxx add cli/piperDetects the agents on your machine, shows the access level and pins the version.
Other ways to install
Access and checks
How we reviewExtended accessThe tool runs code or makes network calls and only reads external data.
- The CLI downloads models and creates audio files; playback mode emits sound on the device.
This is third-party code. Review the repository files before installing.
What it does
Piper is a CLI for turning text into speech on your computer. Users download a voice model, supply text and receive an audio file or playback through ffplay. The built-in espeak-ng prepares word pronunciation for synthesis. OHF-Voice continues the earlier rhasspy Piper project, whose repository explicitly points to the new one. Each voice consists of an ONNX model and a configuration file. For repeated requests, the author recommends the HTTP server instead of reloading the model for every CLI invocation.
Who it is for. Voice-interface developers and teaching authors who need local speech from prepared text.
Good fit when
- An author turns a short instruction into WAV without cloud TTS.
- A developer selects a voice for a local assistant.
- A teaching project reads text in Russian or another supported language.
- A team moves frequent synthesis requests from separate CLI calls to an HTTP server.
Not a fit when
- An editor wants to clone a person’s voice from a supplied recording.
- A user downloads an ONNX file without its matching voice configuration.
- A commercial project plans to use a voice without checking its separate license.
Example request
Render this short instruction to WAV with Piper, showing the chosen voice and its MODEL_CARD link first.Limitations
Each voice requires both an `.onnx` model and its matching `.onnx.json` configuration. Voice models may carry their own licensing restrictions, documented in each `MODEL_CARD`. A separate CLI invocation reloads the model, so the author recommends the HTTP server for repeated synthesis. CUDA acceleration through `--cuda` requires `onnxruntime-gpu`; install it separately before requesting synthesis on a supported GPU.
How to disable. Stop Piper or its HTTP server and disconnect calls from your application. Voice models and generated WAV files remain on disk.
FAQ
How do I install current Piper?
Install `pip install piper-tts`, then download a voice and generate a test WAV. This is the current OHF-Voice package; the old rhasspy/piper repository points here for continued development.
How do I find and download a voice?
`python3 -m piper.download_voices` lists voices. The sample download is `python3 -m piper.download_voices en_US-lessac-medium`, writing files to the current directory.
How do I generate a first WAV?
After downloading the sample voice, run `python3 -m piper -m en_US-lessac-medium -f test.wav -- 'This is a test.'`. It writes that phrase to `test.wav`.
Are Russian voices available?
Yes. Voices lists `ru_RU`. Select a specific voice through the downloader and read its `MODEL_CARD` before using it in a project.
How do I choose the voice directory?
Downloading and synthesis accept `--data-dir <DIR>`. Substitute your directory and use it in both steps so the CLI finds the downloaded files.
Discussion
No questions or reviews yet.
Related
CLI toolsWhisper is a CLI for audio transcription, with model sizes for different hardware and output in the original spoken language.
Local speech synthesis for agents: a command and a skill voice text, clone a voice from a recording and render scores to audio, with no cloud
A text-to-speech MCP server that turns text and multi-voice scripts into audio through ElevenLabs and keeps a job history.
whisper.cpp is an audio CLI without a Python environment, compiled for a selected processor or hardware accelerator.
A local MCP server for voice work on macOS: voice cloning, speech synthesis and audiobooks from PDF on an agent's command
A skill with two scripts: text to an .ogg voice message and voice to text through Yandex SpeechKit, with ten voices and three emotions
README in short
The README links to CLI, HTTP server and programming APIs. The voice guide lists languages and model files. Earlier binary releases are separate from the current pip installation.