Skip to main content

​
Fish Audio Python SDK

To provide convenient Python program integration for https://docs.fish.audio.

​
Install

pip install fish-audio-sdk

​
Usage

The new SDK uses the fishaudio module: 
from fishaudio import FishAudio

client = FishAudio(api_key="your_api_key")
You can customize the base URL: 
from fishaudio import FishAudio

client = FishAudio(api_key="your_api_key", base_url="https://your-proxy-domain")

​
Legacy SDK

The legacy SDK uses the fish_audio_sdk module. Initialize a Session to use APIs. All APIs have synchronous and asynchronous versions. If you want to use the asynchronous version of the API, you only need to rewrite the original session.api_call(...) to session.api_call.awaitable(...). 
from fish_audio_sdk import Session

session = Session("your_api_key")
Sometimes, you may need to change our endpoint to another address. You can use 
from fish_audio_sdk import Session

session = Session("your_api_key", base_url="https://your-proxy-domain")

​
Text to speech

from fish_audio_sdk import Session, TTSRequest

session = Session("your_api_key")

with open("r.mp3", "wb") as f:
    for chunk in session.tts(TTSRequest(text="Hello, world!")):
        f.write(chunk)
Or use async version: 
import asyncio
import aiofiles

from fish_audio_sdk import Session, TTSRequest

session = Session("your_api_key")


async def main():
    async with aiofiles.open("r.mp3", "wb") as f:
        async for chunk in session.tts.awaitable(
            TTSRequest(text="Hello, world!"),
        ):
            await f.write(chunk)


asyncio.run(main())

​
Reference Audio

from fish_audio_sdk import TTSRequest

TTSRequest(
    text="Hello, world!",
    reference_id="your_model_id",
)
Or just use ReferenceAudio in TTSRequest: 
from fish_audio_sdk import TTSRequest, ReferenceAudio

TTSRequest(
    text="Hello, world!",
    references=[
        ReferenceAudio(
            audio=audio_file.read(),
            text="reference audio text",
        )
    ],
)

​
List models

models = session.list_models()
print(models)
Or use async version: 
import asyncio


async def main():
    models = await session.list_models.awaitable()
    print(models)


asyncio.run(main())

​
Get a model info by id

model = session.get_model("your_model_id")
print(model)
Or use async version: 
import asyncio


async def main():
    model = await session.get_model.awaitable("your_model_id")
    print(model)


asyncio.run(main())

​
Create a model

model = session.create_model(
    title="test",
    description="test",
    voices=[voice_file.read(), other_voice_file.read()],
    cover_image=image_file.read(),
)
print(model)
Or use async version: 
import asyncio


async def main():
    model = await session.create_model.awaitable(
        title="test",
        description="test",
        voices=[voice_file.read(), other_voice_file.read()],
        cover_image=image_file.read(),
    )
    print(model)


asyncio.run(main())

​
Delete a model

session.delete_model("your_model_id")
Or use async version: 
import asyncio


async def main():
    await session.delete_model.awaitable("your_model_id")


asyncio.run(main())