Deepgram Deepgram

ASR Reinvented

The fastest, most accurate, most reliable, most scalable speech transcription, with understanding — rebuilt just for enterprise.

Say goodbye to flatlining accuracy.

We’ve reinvented Automatic Speech Recognition (ASR) with 100% deep learning that allows companies to continuously improve accuracy. Stop waiting for the big tech players to improve their software and forcing your developers to manually boost accuracy with keywords in every API call. Start training your speech model and reaping the benefits in weeks, not months or years.

Deepgram

The Speech Platform for Enterprise

The one platform you need to label, train and deploy speech recognition models at scale. Convert 100% of audio data to start surfacing insights and automating actions.

1. LABEL

Leverage the latest data curation and labeling tools to improve model accuracy

Learn More

2. TRAIN

Get started quickly with Beginner models. See upwards of 90% accuracy using training.

Learn More

3. DEPLOY

Access AI speech recognition models via APIs without deep data science expertise.

Learn More

4. MANAGE

Upload audio, label data, train and deploy from one place. Monitor usage and request API keys.

Create Account

Designed for Developers

The Easiest and Most Reliable Speech-to-Text API Ever

Use one API to label, train and deploy AI speech models and transcriptions at scale. No need to splice your audio file into shorter durations, decide between synchronous or async, real-time or batch. Deepgram is one API that easily integrates accurate transcripts into your application or existing data processing pipelines.

View Docs

curl \
    -X POST \
    -u USERNAME:PASSWORD \
    -H "Content-Type: application/json" \
    -d '{"url":
"https://www.deepgram.com/examples/interview_speech-analytics.wav"}'
\
  https://brain.deepgram.com/v2/listen?punctuate=true
import base64
import json
import urllib.request

url = 'https://brain.deepgram.com/v2/listen'
username = 'USERNAME'
password = 'PASSWORD'

headers = {}
headers['Authorization'] = 'Basic {}'.format(
base64.b64encode('{}:{}'.format(username,
password).encode('utf-8')).decode('utf-8')
)
headers['Content-Type'] = 'application/json'

data = { 'url': 'https://www.deepgram.com/examples/interview_speech-analytics.wav' }

req = urllib.request.Request( url, method='POST', headers=headers, data=json.dumps(data).encode('utf-8') )
resp = urllib.request.urlopen(req)
const axios = require('axios');

let url = 'https://brain.deepgram.com/v2/listen';
let username = 'USERNAME';
let password = 'PASSWORD';
let audio = 'https://www.deepgram.com/examples/interview_speech-analytics.wav';

axios({
  method: 'post',
  url: url,
  auth: {
    username username,
    password: password
},
headers: {
  'Content-Type': 'application/json'
},
data: {
    url: audio
}
})
.then(response => {
  console.log(response.data);
})
.catch(error => {
  console.log('Error happened!: ' + error);
});
curl \
    -X POST \
    -u USERNAME:PASSWORD \
    -H "Content-Type: application/json" \
    -d '{"url":
"https://www.deepgram.com/examples/interview_speech-analytics.wav"}'
\
  https://brain.deepgram.com/v2/listen?punctuate=true
import base64
import json
import urllib.request

url = 'https://brain.deepgram.com/v2/listen'
username = 'USERNAME'
password = 'PASSWORD'

headers = {}
headers['Authorization'] = 'Basic {}'.format(
base64.b64encode('{}:{}'.format(username,
password).encode('utf-8')).decode('utf-8')
)
headers['Content-Type'] = 'application/json'

data = { 'url': 'https://www.deepgram.com/examples/interview_speech-analytics.wav' }

req = urllib.request.Request( url, method='POST', headers=headers, data=json.dumps(data).encode('utf-8') )
resp = urllib.request.urlopen(req)
const axios = require('axios');

let url = 'https://brain.deepgram.com/v2/listen';
let username = 'USERNAME';
let password = 'PASSWORD';
let audio = 'https://www.deepgram.com/examples/interview_speech-analytics.wav';

axios({
  method: 'post',
  url: url,
  auth: {
    username username,
    password: password
},
headers: {
  'Content-Type': 'application/json'
},
data: {
    url: audio
}
})
.then(response => {
  console.log(response.data);
})
.catch(error => {
  console.log('Error happened!: ' + error);
});

Deepgram engines have outperformed any of the others that we have tried or looked into. Our accuracy levels are greater than 90% on virtually everything that we do.”

Dennis Evanson

Compliance and QA
Randall Reilly

95%

Increased audit-confirmed results to levels exceeding 95% accuracy.

Read Their Story

Being able to rely on Deepgram transcription, both on the front and back end of the call is paramount to accurate emotion detection for our Call Center Customers.”

Adam Settle

VP of Product
Sharpen

100%

Captured 100% of our Call Center audio.

Watch Webinar

There could be hundreds of issues a customer is calling in about. Add to this complexity there is a distribution of words, specific to each of our customer’s brands. We couldn’t get these words right using Google, Amazon or Speechmatics, and are thrilled to finally reach our accuracy goal with Deepgram.”

Arjun Maheswaran

CTO
Agara

In a head to head test, Deepgram model training yielded a lower WER.

Learn More

Deepgram is doing groundbreaking work in the speech analytics field, and we are delighted to be working closely with them. Their world class GPU-accelerated speech recognition enables faster, more accurate natural language processing that will make an important impact on a range of industries.”

Jeff Herbst

VP of Business Development
NVIDIA

11

Powered by the NVIDIA GPU architecture and 11 Deep Learning patents, Deepgram is the most cost efficient ASR.

View Blog

Google dumped out a big, disgusting JSON file with quality that wasn’t good enough. Deepgram was first accuracy-wise and produced, by far, the easiest transcriptions to work with.”

Klint Kanopka

PhD Student Research
Stanford’s Graduate School of Education

#1

Deepgram was first accuracy-wise when compared to Google.

Read Their Story

Why deepgram?

100% Deep Learning

Like you, we weren’t satisfied with the way automatic transcription worked — or more accurately, didn’t work — so we rebuilt it. We scrapped the traditional methods and opted for complete deep learning models that deliver speech-to-text on a level that the big tech players can’t touch.

Unlike other ASRs, Deepgram isn’t stagnant. It auto-learns from your data. Take it for a spin. Train a model to recognize the data you care about. It’ll continuously improve, delivering even greater accuracy and understanding over time.

Why it Matters?
  • Maximum Accuracy

    Over 90% Accuracy. The only AI speech platform that trains on your unique audio.

  • Accelerated Time to Value

    120x Faster. Transcribe hour long recordings in seconds. Get superior speech models in weeks, not months or years.

  • Continuous Learning

    11 Patents for deep neural networking, which allow models to increase accuracy at unprecedented speed and cost.

  • Resilient Operations

    State-of-the-art infrastructure supports 100s of models across our customer base. Process billions of words per day with 99.9% availability.

  • Future-Proof Foundation

    Flexible and built for change. Train models and deploy anywhere – on premises or in the cloud.

Create a future with voice

The Foundation for a AI-Enabled Enterprise

Across industries, Deepgram makes converting speech to text simpler than ever before. Our GPU enabled, patented deep learning approach to speech recognition allows us to process billions of hours every year. Unlock your audio and start transforming your business.

See Overview

Start getting more value out of your speech data.

Contact Us

Contact Us

Speech recognition is hard. We'll make it easy.