Resources · Make integration

Deepgram Make integrationAutomate Deepgram with Make.

What if every recording your team collects came back as text, without anyone pressing play? The Deepgram Make integration gives you 3 action modules to transcribe audio from a URL and summarize it. This page shows how to connect your API key and build a first scenario.

Verified Trustpilot reviews · AI, automation & growth agency

What you can automate

What can the Deepgram Make integration do for you?

The Deepgram Make integration lets a Make scenario (the automation you build, a chain of steps) send audio to Deepgram and get text back. You pick a module, which is one brick of that chain, point it at a recording, and Make passes the transcript or the summary to whatever tool comes next.

Call recordings that document themselves. A sales team stores call audio somewhere with a public link. Another app hands that link to Make, Transcribe a Prerecorded Audio File from URL turns it into text, and the text lands in the CRM next to the deal.

Podcast show notes from the raw episode. A content team publishes an episode, Summarize an Audio File condenses it, and the summary becomes the starting draft for the episode page. Nobody has to listen twice.

Anything the two modules skip. Deepgram's API does more than these two jobs. Make an API Call reaches any endpoint of that API with the connection you already set up.

Deepgram has no trigger in Make: nothing on the Deepgram side starts a scenario. The recording has to come from another app, a schedule, or a webhook. If you are still weighing tools, the n8n vs Make comparison lays out both, and the Make review covers the editor itself. For a hands-on start, see Make training.

Connect

How do you connect Deepgram to Make?

  1. 01

    Create an API key in Deepgram

    In Deepgram, open the Projects dropdown at the top left and choose your project. Go to Settings, then the API Keys view, click Create a New API Key, fill in the settings and confirm with Create Key.

  2. 02

    Copy the key right away

    Deepgram shows the new key in a modal window. Copy it before closing that window and keep it somewhere safe, because Make will ask for it in the next step. You need an active Deepgram account for any of this to work.

  3. 03

    Paste it into Make

    In your scenario, add any Deepgram module and click Create a connection. Give it a Connection name if you like, paste the key, and save. That connection is your Deepgram account linked to Make, and every Deepgram module reuses it.

First scenario

Your first scenario with Deepgram

GoalWhen another app hands Make a link to an audio file, Deepgram transcribes it and then summarizes it.

  1. 01

    Create the scenario

    On the Scenarios page, create a new scenario and click the +. Pick the app that holds your recordings as the first module, since Deepgram has no trigger of its own.

  2. 02

    Add the transcription module

    Click the + on the right of that first module, search for Deepgram and choose Transcribe a Prerecorded Audio File from URL. Select your connection, or click Create a connection if this is your first Deepgram module.

  3. 03

    Map the audio link

    In the module fields, map the audio URL coming from the first module. Each bundle, meaning one item passed along the chain, carries one recording, so each one gets its own transcript.

  4. 04

    Add the summary

    Click the + again and add Summarize an Audio File. Point it at the same recording, checking in its fields what it asks for, and you get a short version next to the full transcript.

  5. 05

    Test with Run once

    Click Run once. Make runs the scenario one time, and you can open each module to read the bundles it received and check the transcript before going further.

  6. 06

    Schedule and switch on

    Open the clock on the first module to set the schedule, every 15 minutes by default, then switch the scenario on. From then on, new recordings get processed without you.

Modules

What each Deepgram module does

Deepgram gives you 3 modules. For each one: what it does for you, when to reach for it, and what to watch out for.

Deepgram1

Make an API Call

Action

In MakePerforms an arbitrary authorized API call.

This module sends a request of your choice to Deepgram's API, signed with your existing connection, so you reach features the two ready-made modules leave out.

When to use it
you need a Deepgram capability that neither the transcription nor the summary module offers, and you know which endpoint handles it.
Watch out
nothing here is prefilled for you, so you describe the request yourself and read the raw answer in Run once before mapping it further.
Deepgram2

Summarize an Audio File

Action

In MakeSummarizes the content of an audio file with Deepgram's API.

You get a condensed version of what was said in a recording, handy when people need the gist of a meeting or an episode rather than every word.

When to use it
a manager wants a short recap posted in the team chat after each recorded client call, without reading a long transcript.
Watch out
a summary drops detail by design, so pair it with the transcription module whenever someone may need the exact wording later.
Deepgram3

Transcribe a Prerecorded Audio File from URL

Action

In MakeTranscribes URL audio files with Deepgram's prerecorded API.

Give this module a link to an existing recording and it returns the spoken words as text, ready to store, search or pass to the next module.

When to use it
interviews, voicemails or webinars already sit online and you want a searchable written copy next to each one.
Watch out
the module works from a URL, so the file has to be reachable at that address. A link that only opens after a login on your side will most likely fail, so test one in Run once.
Need help

Need help automating Deepgram with Make?

A person reads every message.

FAQ

Deepgram and Make: common questions

01Is the Deepgram Make integration free?
Yes, on the Make side. Deepgram is a standard app, so it is available from the Free plan, which allows 2 active scenarios, a minimum of 15 minutes between scheduled runs, 5 minutes of execution per run, files up to 5 MB and 512 MB of data transfer. Paid plans lower the interval to 1 minute and lift the active scenario limit. Deepgram itself needs an active account with them, and its own pricing is set by Deepgram, so check their pricing page before you plan volumes.
02What does the Deepgram Make integration need to connect?
An active Deepgram account and an API key. You create the key in Deepgram by choosing your project in the Projects dropdown, opening Settings, then the API Keys view, and clicking Create a New API Key. Copy the key from the modal that appears. In Make, add a Deepgram module, click Create a connection, optionally name it, and paste the key. The same connection then serves all three Deepgram modules in any scenario you build.
03Can a Deepgram scenario start in real time?
Not from Deepgram, because the app has no trigger module in Make. The scenario has to start elsewhere. You can open it with a trigger from the app where your recordings arrive, or set a schedule on the first module, which runs every 15 minutes by default on a new scenario. A webhook from Make is another option when a tool can call a URL. Whatever starts the flow, the Deepgram modules run right after it in the same scenario.
04What if a Deepgram module is missing in Make?
Use Make an API Call. It calls any endpoint of the Deepgram API with the connection you already have, so a feature outside transcription and summary stays within reach. Keep in mind that the Make documentation page for Deepgram says it was generated with AI from Deepgram's own documentation and may contain mistakes. When a detail matters, confirm it in the Deepgram documentation or with Deepgram directly before building around it.
05Make or n8n for Deepgram: which one fits?
It depends on the tool your team already works in. An app like Deepgram often has an n8n node as well. In Make you get three ready modules, the transcription from a URL, the summary and the API call, inside a visual editor where you test with Run once. If your other automations already live in one of the two tools, keeping Deepgram there saves you a second set of connections to maintain.