Implementation:Elevenlabs Elevenlabs python Audio File Pattern

Knowledge Sources	ElevenLabs Python
Domains	Voice_Cloning, Data_Preparation
Last Updated	2026-02-15 00:00 GMT

Overview

User-defined pattern for preparing audio files as core.File objects for the voice cloning API.

Description

This is a Pattern Doc. The core.File type in the ElevenLabs SDK accepts three forms:

• A file-like object (e.g., open("file.mp3", "rb"))
• A tuple of (filename, file-like)
• A tuple of (filename, file-like, content_type)

Users prepare a list of these objects to pass to client.voices.ivc.create(files=...).

Usage

Open audio files in binary mode and collect them into a list before calling the voice cloning API.

Interface Specification

import typing

# core.File type (simplified):
File = typing.Union[
    typing.IO[bytes],                                    # File-like object
    typing.Tuple[str, typing.IO[bytes]],                 # (filename, file-like)
    typing.Tuple[str, typing.IO[bytes], str],            # (filename, file-like, content_type)
]

# Usage:
files: typing.List[File] = [
    open("sample1.mp3", "rb"),
    ("sample2.wav", open("sample2.wav", "rb"), "audio/wav"),
]

I/O Contract

Inputs

Name	Type	Required	Description
audio files	file paths on disk	Yes	Audio recordings of the target speaker

Outputs

Name	Type	Description
files	List[core.File]	List of file objects ready for upload to voices.ivc.create()

Example Implementations

Simple File Open

# Open files directly
files = [
    open("voice_sample_1.mp3", "rb"),
    open("voice_sample_2.mp3", "rb"),
]

voice = client.voices.ivc.create(name="My Voice", files=files)

With Context Manager

from pathlib import Path

sample_dir = Path("./voice_samples")
sample_files = list(sample_dir.glob("*.mp3"))

# Open all samples
files = [open(f, "rb") for f in sample_files]
try:
    voice = client.voices.ivc.create(
        name="My Voice",
        files=files,
        remove_background_noise=True,
    )
finally:
    for f in files:
        f.close()

Related Pages

Implements Principle

• Principle:Elevenlabs_Elevenlabs_python_Audio_Sample_Preparation