Orpheus

Orpheus#

Orpheus is a 3B parameter text-to-speech model developed by Canopy Labs, using a SNAC audio tokenizer for high-quality speech synthesis.

Quickstart#

Start the server:

python -m vox_serve.launch --model canopylabs/orpheus-3b-0.1-ft --port 8000

Generate speech:

import requests

response = requests.post("http://localhost:8000/generate", json={
    "text": "Hello, this is Orpheus speaking!",
    "voice": "tara"
})

with open("output.wav", "wb") as f:
    f.write(response.content)

Available Voices#

Orpheus supports 8 built-in voices:

Voice

Description

tara

Default female voice

leah

Female voice

jess

Female voice

mia

Female voice

zoe

Female voice

leo

Male voice

dan

Male voice

zac

Male voice

API Parameters#

Parameter

Type

Default

Description

text

string

required

Text to synthesize

voice

string

"tara"

Voice to use for synthesis

streaming

boolean

false

Enable streaming response

Examples#

Basic Usage#

import requests

response = requests.post("http://localhost:8000/generate", json={
    "text": "The quick brown fox jumps over the lazy dog.",
    "voice": "leo"
})

Streaming Audio#

import requests

with requests.post(
    "http://localhost:8000/generate",
    json={"text": "Hello world!", "streaming": True},
    stream=True
) as response:
    with open("output.wav", "wb") as f:
        for chunk in response.iter_content(chunk_size=8192):
            f.write(chunk)

Using curl#

curl -X POST "http://localhost:8000/generate" \
  -H "Content-Type: application/json" \
  -d '{"text": "Hello from Orpheus!", "voice": "tara"}' \
  -o output.wav

CLI Options#

python -m vox_serve.launch \
  --model canopylabs/orpheus-3b-0.1-ft \
  --port 8000 \
  --temperature 0.7 \
  --top_p 0.9 \
  --repetition_penalty 1.1
Contents