Synchronous Speech to Text — ElevenLabs Documentation

In this tutorial, you’ll learn how to transcribe audio with the ElevenLabs SDK synchronously.

Requirements

An ElevenLabs account with an API key.
Python or Node installed on your machine

Setup

Installing the SDK

Before you begin, make sure you have installed the ElevenLabs SDK.

1 pip install elevenlabs

Additionally, install necessary packages to manage your environmental variables:

1 pip install python-dotenv

Next, create a .env file in your project directory and fill it with your credentials like so:

.env

1 ELEVENLABS_API_KEY=your_elevenlabs_api_key_here

To convert the audio of a file to text, we’ll use the convert method of the ElevenLabs SDK.

Convert speech to text

1 from dotenv import load_dotenv
2 from io import BytesIO
3 import requests
4 from elevenlabs.client import ElevenLabs
5 
6 load_dotenv()
7 
8 client = ElevenLabs(
9     api_key=os.getenv("ELEVENLABS_API_KEY"),
10 )
11 
12 audio_url = (
13     "https://storage.googleapis.com/eleven-public-cdn/audio/marketing/nicole.mp3"
14 )
15 response = requests.get(audio_url)
16 audio_data = BytesIO(response.content)
17 
18 transcription = client.speech_to_text.convert(
19     file=audio_data,
20     model_id="scribe_v1", # Model to use, for now only "scribe_v1" is supported
21     tag_audio_events=True, # Tag audio events like laughter, applause, etc.
22     language_code="eng", # Language of the audio file. If set to None, the model will detect the language automatically.
23     diarize=True, # Whether to annotate who is speaking
24 )
25 
26 print(transcription.text)

Run the script with:

1 python speech_to_text.py

1	from dotenv import load_dotenv
2	from io import BytesIO
3	import requests
4	from elevenlabs.client import ElevenLabs
5
6	load_dotenv()
7
8	client = ElevenLabs(
9	api_key=os.getenv("ELEVENLABS_API_KEY"),
10	)
11
12	audio_url = (
13	"https://storage.googleapis.com/eleven-public-cdn/audio/marketing/nicole.mp3"
14	)
15	response = requests.get(audio_url)
16	audio_data = BytesIO(response.content)
17
18	transcription = client.speech_to_text.convert(
19	file=audio_data,
20	model_id="scribe_v1", # Model to use, for now only "scribe_v1" is supported
21	tag_audio_events=True, # Tag audio events like laughter, applause, etc.
22	language_code="eng", # Language of the audio file. If set to None, the model will detect the language automatically.
23	diarize=True, # Whether to annotate who is speaking
24	)
25
26	print(transcription.text)