Fish Audio Plugin

A high-quality Text-to-Speech (TTS) and Speech-to-Text (STT) plugin for Vision Agents that uses the Fish Audio API.

Installation

uv add "vision-agents[fish]"
# or directly
uv add vision-agents-plugins-fish

Usage

Text-to-Speech (TTS)

from vision_agents.plugins.fish import TTS
from getstream.video.rtc.audio_track import AudioStreamTrack

# Initialize with API key from environment variable
tts = TTS()

# Or specify API key directly
tts = TTS(api_key="your_fish_audio_api_key")

# Create an audio track to output speech
track = AudioStreamTrack(framerate=16000)
tts.set_output_track(track)

# Register event handlers
@tts.events.subscribe
async def on_audio(event):
    print(f"Received audio chunk: {len(event.audio_data)} bytes")

# Send text to be converted to speech
async for chunk in tts.send_iter("Hello, this is a test of the Fish Audio text-to-speech plugin."):
    pass

Speech-to-Text (STT)

from vision_agents.plugins.fish import STT
from getstream.video.rtc.track_util import PcmData

# Initialize with API key from environment variable
stt = STT()

# Or specify API key directly and language
stt = STT(api_key="your_fish_audio_api_key", language="en")

# Register event handlers
@stt.events.subscribe
async def on_transcript(event):
    print(f"Transcript: {event.text}")

# Process audio data
pcm_data = PcmData(samples=audio_samples, sample_rate=16000)
await stt.process_audio(pcm_data)

Configuration Options

TTS Options

• api_key: Fish Audio API key (default: reads from FISH_API_KEY environment variable)
• reference_id: Optional reference voice ID to use for synthesis
• base_url: Optional custom API endpoint (default: uses Fish Audio's default endpoint)
• client: Optionally pass in your own instance of the Fish Audio Session

STT Options

• api_key: Fish Audio API key (default: reads from FISH_API_KEY environment variable)
• language: Language code for transcription (e.g., "en", "zh"). If None, automatic language detection will be used
• ignore_timestamps: Skip timestamp processing for faster results (default: False)
• sample_rate: Sample rate of the audio in Hz (default: 16000)
• base_url: Optional custom API endpoint
• client: Optionally pass in your own instance of the Fish Audio Session

Reference Audio

Fish Audio supports using reference audio for voice cloning:

from vision_agents.plugins.fish import TTS

# Using a reference voice ID
tts = TTS(reference_id="your_reference_voice_id")

# Or pass reference audio dynamically when sending text
# (See Fish Audio SDK documentation for advanced usage)

Supported Languages (STT)

Fish Audio STT supports multiple languages with automatic detection. Common language codes include:

• en - English
• zh - Chinese
• es - Spanish
• fr - French
• de - German
• ja - Japanese
• ko - Korean
• pt - Portuguese

For automatic language detection, set language=None (default).

Supported Audio Formats (STT)

The STT implementation accepts PCM audio data and converts it to WAV format internally. Supported configurations:

• Maximum audio size: 100MB
• Maximum duration: 60 minutes
• Sample rate: 16kHz or higher recommended
• Format: Mono, 16-bit PCM

Requirements

• Python 3.10+
• fish-audio-sdk>=2025.4.2

Getting Your API Key

1. Sign up for a Fish Audio account at https://fish.audio
2. Navigate to the API Keys section in your dashboard
3. Create a new API key
4. Set the FISH_API_KEY environment variable or pass it directly to the plugin