Speech Recognition

Deepgram

Learn how to use Deepgram with Composio

Overview

SLUG: DEEPGRAM

Description

Deepgram provides AI-powered speech recognition and understanding services, offering APIs for real-time and pre-recorded audio transcription, text-to-speech, and audio intelligence.

Authentication Details

generic_api_key
stringRequired

Connecting to Deepgram

Create an auth config

Use the dashboard to create an auth config for the Deepgram toolkit. This allows you to connect multiple Deepgram accounts to Composio for agents to use.

1

Select App

Navigate to Deepgram.

2

Configure Auth Config Settings

Select among the supported auth schemes of and configure them here.

3

Create and Get auth config ID

Click “Create Deepgram Auth Config”. After creation, copy the displayed ID starting with ac_. This is your auth config ID. This is not a sensitive ID — you can save it in environment variables or a database. This ID will be used to create connections to the toolkit for a given user.

Connect Your Account

Using API Key

1from composio import Composio
2
3# Replace these with your actual values
4deepgram_auth_config_id = "ac_YOUR_DEEPGRAM_CONFIG_ID" # Auth config ID created above
5user_id = "0000-0000-0000"  # UUID from database/app
6
7composio = Composio()
8
9def authenticate_toolkit(user_id: str, auth_config_id: str):
10    # Replace this with a method to retrieve an API key from the user.
11    # Or supply your own.
12    user_api_key = input("[!] Enter API key")
13
14    connection_request = composio.connected_accounts.initiate(
15        user_id=user_id,
16        auth_config_id=auth_config_id,
17        config={"auth_scheme": "API_KEY", "val": user_api_key}
18    )
19
20    # API Key authentication is immediate - no redirect needed
21    print(f"Successfully connected Deepgram for user {user_id}")
22    print(f"Connection status: {connection_request.status}")
23    
24    return connection_request.id
25
26
27connection_id = authenticate_toolkit(user_id, deepgram_auth_config_id)
28
29# You can verify the connection using:
30connected_account = composio.connected_accounts.get(connection_id)
31print(f"Connected account: {connected_account}")

Tools

Executing tools

To prototype you can execute some tools to see the responses and working on the Deepgram toolkit’s playground

Python
1from composio import Composio
2from openai import OpenAI
3import json
4
5openai = OpenAI()
6composio = Composio()
7
8# User ID must be a valid UUID format
9user_id = "0000-0000-0000"  # Replace with actual user UUID from your database
10
11tools = composio.tools.get(user_id=user_id, toolkits=["DEEPGRAM"])
12
13print("[!] Tools:")
14print(json.dumps(tools))
15
16def invoke_llm(task = "What can you do?"):
17    completion = openai.chat.completions.create(
18        model="gpt-4o",
19        messages=[
20            {
21                "role": "user",
22                "content": task,  # Your task here!
23            },
24        ],
25        tools=tools,
26    )
27
28    # Handle Result from tool call
29    result = composio.provider.handle_tool_calls(user_id=user_id, response=completion)
30    print(f"[!] Completion: {completion}")
31    print(f"[!] Tool call result: {result}")
32
33invoke_llm()

Tool List

Tool Name: Get Public Models

Description

Tool to retrieve metadata on all the latest public deepgram speech-to-text models. use when you need to list available models; set include outdated to true to include deprecated versions.

Action Parameters

include_outdated
boolean

Action Response

data
objectRequired
error
string
successful
booleanRequired

Tool Name: List Deepgram Projects

Description

Tool to list all deepgram projects. use after authenticating with your api key.

Action Parameters

Action Response

data
objectRequired
error
string
successful
booleanRequired

Tool Name: Get Project Usage Summary

Description

Tool to retrieve a summary of usage data for a specified deepgram project. use when you need high-level metrics (submitted, processed, billable durations, etc.) optionally filtered by time window, model, accessor, or tag.

Action Parameters

accessor_id
string
end
string
model
string
project_id
stringRequired
start
string
tag
string

Action Response

data
objectRequired
error
string
successful
booleanRequired

Tool Name: Get Public TTS Models

Description

Tool to fetch metadata about all latest public tts voice models. use when you need to list available deepgram tts voices.

Action Parameters

include_outdated
boolean

Action Response

data
objectRequired
error
string
successful
booleanRequired

Tool Name: List Project Scopes

Description

Tool to list all scopes for a specified deepgram project. use when you need to retrieve all permission scopes for a project.

Action Parameters

project_id
stringRequired

Action Response

data
objectRequired
error
string
successful
booleanRequired

Tool Name: Transcribe Pre-recorded Audio

Description

Tool to transcribe pre-recorded audio files into text. use when you need to convert an audio file at a url to text after confirming the file url is publicly accessible.

Action Parameters

alternatives
integer
audio_url
stringRequired
content_type
stringRequired
detect_language
boolean
diarize
boolean
interim_results
boolean
keywords
array
language
string
model
string
punctuate
boolean
smart_format
boolean
tier
string
version
string

Action Response

data
objectRequired
error
string
successful
booleanRequired

Tool Name: Summarize Audio

Description

Tool to generate concise summaries from audio content. use when you need a quick summary of audio after confirming the file url is publicly accessible.

Action Parameters

audio_url
stringRequired
content_type
stringRequired
language
string
model
string
punctuate
boolean
smart_format
boolean
tier
string

Action Response

data
objectRequired
error
string
successful
booleanRequired

Tool Name: Text-to-Speech (REST)

Description

Tool to convert text into natural-sounding speech. use when you need tts audio from text inputs.

Action Parameters

container
string
encoding
string
language
string
model
string
pitch
numberDefaults to 1
sample_rate
integer
speed
numberDefaults to 1
text
stringRequired
version
string
voice
string

Action Response

data
objectRequired
error
string
successful
booleanRequired

Tool Name: Topic Detection

Description

Tool to transcribe audio and detect topics. use when you need to identify the main topics discussed in an audio recording after confirming the file url is publicly accessible.

Action Parameters

audio_url
stringRequired
content_type
stringRequired
language
string
model
string

Action Response

data
objectRequired
error
string
successful
booleanRequired