Inference Logoinference shell

Fish Audio Drama 3

Fish Audio Drama 3 text to speech. Direct the delivery in plain language, add laughs and sighs, clone a voice from a clip, or voice a dialogue. 80+ languages.

text to speechvoice cloning
run in browser →run via API
run with your agent
# install belt
$curl -fsSL https://cli.inference.sh | sh
# sign in (or set INFERENCE_API_KEY on a headless machine)
$belt login
# view schema & details
$belt app get fish-audio/drama-3
# run
$belt app run fish-audio/drama-3 --input '{"text":"text","voice_ids":["bf322df2096a46f18c579d0baa36f41d"]}'
maker
fish-audio
category
audio
input → output
text → speech
price
$50.00/1M UTF-8 bytes
runs via
api, javascript and python sdk, belt cli, mcp
updated
2026-10-15

api reference

about

fish audio drama 3 text to speech. direct the delivery in plain language, add laughs and sighs, clone a voice from a clip, or voice a dialogue. 80+ languages.

1. calling the api

install the client

the client provides a convenient way to interact with the api.

bash
1pip install inferencesh

setup your api key

set INFERENCE_API_KEY as an environment variable. get your key from settings → api keys.

bash
1export INFERENCE_API_KEY="inf_your_key"

run and get result

submit a request and wait for the final result. best for batch processing or when you don't need progress updates.

python
1import os2from inferencesh import inference34client = inference(api_key=os.environ["INFERENCE_API_KEY"])567result = client.run({8        "app": "fish-audio/drama-3",9        "input": {10            "text": "text",11            "voice_ids": [12                "bf322df2096a46f18c579d0baa36f41d"13            ]14        }15    })1617print(result["output"])

stream live updates

get real-time progress updates as the task runs. ideal for showing progress bars, partial results, or long-running tasks.

python
1import os2from inferencesh import inference34client = inference(api_key=os.environ["INFERENCE_API_KEY"])567# stream=True yields updates as they arrive8for update in client.run({9        "app": "fish-audio/drama-3",10        "input": {11            "text": "text",12            "voice_ids": [13                "bf322df2096a46f18c579d0baa36f41d"14            ]15        }16    }, stream=True):17    if update.get("progress"):18        print(f"progress: {update['progress']}%")19    if update.get("output"):20        print(f"output: {update['output']}")

2. authentication

the api uses api keys for authentication. see the authentication docs for detailed setup instructions.

3. files

file inputs are automatically handled by the sdk. you can pass local paths, urls, or base64 data.

automatic upload

the python sdk automatically detects local file paths and uploads them. urls are passed through as-is.

python
1# local file paths are automatically uploaded2result = client.run({3    "app": "fish-audio/drama-3",4    "input": {5        "image": "/path/to/local/image.png",  # detected & uploaded6        "audio": "https://example.com/audio.mp3",  # url passed through7    }8})

manual upload

you can also upload files manually and use the returned url.

python
1# upload and get a hosted URL2file = client.files.upload("/path/to/file.png")3print(file.uri)  # https://cloud.inference.sh/...

4. webhooks

get notified when a task completes by providing a webhook url. when the task reaches a terminal state (completed, failed, or cancelled), a POST request is sent to your url with the task result.

python
1result = client.run({2    "app": "fish-audio/drama-3",3    "input": {4        "text": "text",5        "voice_ids": [6            "bf322df2096a46f18c579d0baa36f41d"7        ]8    },9    "webhook": "https://your-server.com/webhook"10}, wait=False)

webhook payload

your endpoint receives a JSON POST with the task result:

json
1{2  "event": "task.completed",3  "timestamp": "2024-01-15T10:30:05Z",4  "data": {5    "id": "task_abc123",6    "short_id": "abc123",7    "status": 10,8    "status_text": "completed",9    "output": { ... },10    "created_at": "2024-01-15T10:30:00Z",11    "updated_at": "2024-01-15T10:30:05Z"12  }13}
eventstringtask.completed, task.failed or task.cancelled
timestampstringiso timestamp of the delivery
data.idstringtask id
data.short_idstringshort task id
data.statusnumberterminal status (10=completed, 11=failed, 12=cancelled)
data.status_textstringthe status as a word: completed, failed or cancelled
data.outputobjecttask output (when completed)
data.errorstringerror message (when failed, omitted otherwise)
data.session_idstringsession id (when the task ran in a session, omitted otherwise)
data.created_atstringiso timestamp
data.updated_atstringiso timestamp

5. schema

input

textstring*

what to say, and how. bracketed directions and sounds are performed, not read: [barely holding it together] i'm fine. [small laugh] paired tags shape specific words: <whisper>, <emphasis>, <soft>, <fast>, <slow>, <stress>, <forceful>. for dialogue, mark turns with <|speaker:0|>, <|speaker:1|>, ... matching the order of voice_ids.

minLength:1maxLength:30000
voice_idsarray

fish audio voice ids (from fish.audio). one id for a single voice; several for a dialogue, one per speaker tag in order. empty: the model picks a voice.

example: ["bf322df2096a46f18c579d0baa36f41d"]
reference_audiostring(file)

a clip of the voice to clone: 10 to 30 seconds of clear speech, wav, mp3 or flac. use it instead of voice_ids.

reference_textstring

exact transcript of reference_audio. required with it.

formatstring

audio format. pcm is raw 16-bit mono.

default: "mp3"
options:"mp3""wav""opus""pcm"
sample_rateinteger

sample rate in hz. empty: 44100, or 48000 for opus. mp3 takes 32000 or 44100; opus only 48000.

options:80001600024000320004410048000
speednumber

speaking speed, 0.5 to 2.

default: 1min:0.5max:2
temperaturenumber

sampling temperature, 0 to 1. higher is more varied. empty: fish audio's default.

min:0max:1
top_pnumber

nucleus sampling, 0 to 1. empty: fish audio's default.

min:0max:1
normalizeboolean

normalize numbers and symbols in english and chinese text before speaking.

default: true
latencystring

normal for best quality; balanced and low trade quality for a faster start.

default: "normal"
options:"normal""balanced""low"

output

audiostring(file)*

the generated speech.

other audio apps

Eleven Flash v2.5

elevenlabs/eleven-flash-v2-5

$0.04/1K characters

Eleven Multilingual v2

elevenlabs/eleven-multilingual-v2

$0.08/1K characters

Eleven v3

elevenlabs/eleven-v3

$0.08/1K characters

Eleven v4 Turbo

elevenlabs/eleven-v4-turbo

$0.04/1K characters

Eleven v4

elevenlabs/eleven-v4

$0.08/1K characters

new models

  1. Gemini 4 Argon

    google/gemini-4-argon

    chat$2.00/M input, $10.00/M output

  2. GPT-6 Sol

    openai/gpt-6-sol

    chat$2.00/M input, $0.20/M cached input, $10.00/M output

  3. GPT-6 Luna

    openai/gpt-6-luna

    chat$0.10/M input, $0.01/M cached, $0.50/M output

  4. GPT-6.1 Sol

    openai/gpt-6-1-sol

    chat$2.00/M input, $0.10/M cached, $10.00/M output

  5. Claude Fable 5.1

    anthropic/claude-fable-5-1

    chat$10.00/M input, $50.00/M output, $0.25/M cache read, $12.50/M cache write

ready to run Fish Audio Drama 3?

try in browserbrowse all apps

we use cookies

we use cookies to ensure you get the best experience on our website. for more information on how we use cookies, please see our cookie policy.

by clicking "accept", you agree to our use of cookies.
learn more.