Orpheus Tts: Open-Source Text-to-Speech

Understand the project

Practical overview

When it fits

Use Orpheus TTS when you need model and code resources for building or prototyping text-to-speech workflows, rather than a finished consumer-facing audio app.

Who it is for

Developers adding speech generation to applications.
Teams prototyping English or multilingual text-to-speech experiences.
Researchers and practitioners exploring model fine-tuning or speech synthesis workflows.

Documented uses

Convert text into generated speech.
Experiment with streamed audio generation for speech applications.
Explore English models and a multilingual research release.
Use provided scripts and sample datasets as a starting point for fine-tuning.

Before you choose

Requirements & limits

The repository is public and lists an Apache-2.0 license. It documents local Python inference, Colab notebooks and a separate Baseten hosted inference option.

The repository offers a fine-tuned English model and a pretrained English model, plus a multilingual family described as a research release.
The documented streaming example uses the orpheus-speech package and vLLM; the README notes package-version issues and a workaround.
Hosted streaming through Baseten is a separate deployment path from running the repository's example.
The repository includes data-processing scripts and sample datasets for creating fine-tunes.

Check current requirements in the official documentation. Source-based guidance does not mean OpenViST has run or independently tested this project.

Source-based practical guide

Generate streamed speech with the Python example

Use the repository's streaming inference example to generate speech and save the audio as a WAV file.

Before you start

Clone the Orpheus-TTS repository.
Install the orpheus-speech package; the example uses vLLM for inference.
Use the fine-tuned production model named in the example.

Steps

Clone the repository with the command below.
Documented command
```
git clone https://github.com/canopyai/Orpheus-TTS.git
```
Install the package with the command below.
Documented command
```
pip install orpheus-speech
```
Initialize OrpheusModel with canopylabs/orpheus-tts-0.1-finetune-prod, provide text and a voice such as tara, and call generate_speech.
Write each returned audio chunk to a WAV file using the example's mono, 16-bit, 24 kHz settings.

Expected outcome

A WAV file containing speech generated from the supplied text, with audio returned in chunks.

Guide limitations

The repository notes a vLLM package issue and says some bugs are being resolved; it suggests trying pip install vllm==0.7.3 after installing orpheus-speech.
For reported KV-cache or max_model_len errors, the repository describes importing the local package instead of the installed PyPI version.
The example documents one model and voice; it does not establish performance or compatibility for other environments.

Documentation 1 ↗

COMPARE OPTIONS

Compare options for Model inference and serving

These projects document a common task and share a project role. This is a source-based comparison, not an execution benchmark or a guarantee of interchangeability.

Project	Documented purpose	Platform	Access and requirements
Orpheus Tts	Convert text into generated speech.	Cloud	The repository is public and lists an Apache-2.0 license. It documents local Python inference, Colab notebooks and a separate Baseten hosted inference option.
Edgedict	Decode speech from a YouTube Live video into text.	Cloud

Evidence and transparency

Sources & review

Official repository ↗

Source record details

Source availability: Active
Sources checked: 7 October 2026
Original website title: canopyai/Orpheus-TTS · Official homepage title · retained in its source language

How we review sources Suggest a correction

Orpheus Tts