API Docs for the enterprise endpoint

Open .md

This is a separate endpoint with enterprise features and the ability to handle large files. You can send hours- or even days-long audio to this endpoint and receive highly detailed metadata in response.

To identify music from shorter audio clips and retreive rich metadata, including the detailed composition data Spotify, Apple Music, Musicbrainz and other hold (and not just links), use our main API endpoint. For industry-leading broadcast monitoring, see Streams.

The API is very easy to integrate directly; however, we maintain SDKs for 11 languages.

How the requests are counted

We count requests to the enterprise endpoint as 1 request per 12 seconds of audio.

When you upload a file, our servers consider it to be a group of 12-second-long audio chunks. If you don't want every one of these chunks to be recognized and counted as the requests, there are two parameters that you can use: skip and every.

• skip - the number of 12-second-long audio chunks to skip after the ones that are scanned;
• every - the number of chunks to scan in a row.

So if you want to identify music playing in the first 12 seconds, then skip 48 seconds, then scan 12 seconds again, skip 48 seconds again, etc., these parameters should be skip=4 and every=1. If you want to skip 108 seconds, to have 1 request per 120 seconds of audio, send skip=9 and every=1.

Where to get a token

You can get a token from our API Dashboard.

The first 300 requests are free; our subscription options start at $2 per 1000 requests. We use Stripe for secure payments; we can also accept US domestic and international bank transfers from enterprise customers.

How to send files

There are two ways to send files to the API:

• Provide a URL of the file. Our server will download and recognize music from the file. Send the URL in the url parameter. (You can also send a URL of a web page that contains audio or video instead of the URL of the actual audio file)
• Post the file using multipart/form-data in the usual way the files are uploaded via the browser. Send the file in the file parameter, by POST. This method is useful if the file is not available by a URL.

You can try to send a file from a browser here (or here for csv results instead of json).

post

Send a file

https://enterprise.audd.io/

Request

Form Data Parameters

api_token

REQUIRED

string

The token received from the Dashboard

url

POSSIBLY REQUIRED

string

The URL of the file for the recognition

file

POSSIBLY REQUIRED

binary

The file for the recognition

accurate_offsets

OPTIONAL

string

Set to "true" to get accurate start and end offsets of the recognized tracks

skip

OPTIONAL

integer

How many 12 seconds chunks are skipped after the ones that are recognized

every

OPTIONAL

integer

How many chunks should be recognized in a row

limit

OPTIONAL

integer

The upper bound for the number of chunks the server will recognize

skip_first_seconds

OPTIONAL

integer

How many seconds of the file the server should skip at the beginning before starting the music recognition (e.g., the start time of the first chunk). Should not be used together with use_timecode

use_timecode

OPTIONAL

string

Set to "true" if you want the server to use the time represented in "t", "time_continue", or "start" parameters of the URL, if one exists, as the skip_first_seconds. Should not be used together with skip_first_seconds

Response

Here's an example of a response you can get from the recognition of this hour-long mix. It's a real response (shown here as its first three of 63 chunks), and we replaced the duplicate song results with '...' so it's a bit easier to scroll.

200: OK

{
  "status": "success",
  "result": [
    {
      "songs": [
        {
          "score": 100,
          "artist": "Ali Angel",
          "title": "Tou patou II",
          "album": "Notre histoire",
          "release_date": "2002-06-05",
          "label": "BELIEVE - Pastel Productions",
          "timecode": "00:11",
          "isrc": "FR6V81099777",
          "upc": "3760009882294",
          "song_link": "https://lis.tn/TouPatouII",
          "start_offset": 1,
          "end_offset": 8680
        },
        {
          "score": 100,
          "artist": "Future Shock",
          "title": "Dance Floor",
          "album": "Beyond Forever",
          "release_date": "2020-03-20",
          "label": "Sportn' Life Music Group",
          "timecode": "00:11",
          "isrc": "QM4TX2035991",
          "upc": "194491888015",
          "song_link": "https://lis.tn/uxRpGN",
          "start_offset": 1,
          "end_offset": 8840
        },
        {
          "score": 100,
          "artist": "Anthony Acosta",
          "title": "The Way She Moves",
          "album": "The Way She Moves",
          "release_date": "2008-06-01",
          "label": "2254767 Records DK",
          "timecode": "00:33",
          "isrc": "QZHN32178577",
          "upc": "196164729410",
          "song_link": "https://lis.tn/HgiZkK",
          "start_offset": 1,
          "end_offset": 7820
        }
      ],
      "offset": "00:00"
    },
    {
      "songs": [
        {
          "score": 100,
          "artist": "Barely Alive",
          "title": "Keyboard Killer",
          "album": "Lost in the Internet EP",
          "release_date": "2014-03-10",
          "label": "Disciple Recordings",
          "timecode": "00:41",
          "isrc": "FR4GL1069639",
          "upc": "3610154900148",
          "song_link": "https://lis.tn/KeyboardKiller",
          "start_offset": 1,
          "end_offset": 9660
        }
      ],
      "offset": "00:48"
    },
    {
      "songs": [
        {
          "score": 100,
          "artist": "Barely Alive",
          "title": "Keyboard Killer",
          "album": "Lost in the Internet EP",
          "release_date": "2014-03-10",
          "label": "Disciple Recordings",
          "timecode": "01:31",
          "isrc": "FR4GL1069639",
          "upc": "3610154900148",
          "song_link": "https://lis.tn/KeyboardKiller",
          "start_offset": 1,
          "end_offset": 8440
        }
      ],
      "offset": "01:36"
    }
  ],
  "execution_time": "38.335355495s"
}

Code examples

Send a file URL

Python

Plain HTTP, no SDK:

curl https://enterprise.audd.io/ \
    -F api_token='your api token' \
    -F url='https://audd.tech/djatwork_example.mp3' \
    -F accurate_offsets='true' \
    -F skip='3' \
    -F every='1'

Node

With the SDK:

pip install audd

from audd import AudD

audd = AudD("your-api-token")
matches = audd.recognize_enterprise("https://example.com/full-show.mp3", limit=10)
for m in matches:
    print(f"{m.timecode}  {m.artist} — {m.title}")

Full Python reference →

Or send the request yourself, without the SDK:

import requests
data = {
    'api_token': 'your api token',
    'url': 'https://audd.tech/djatwork_example.mp3',
    'accurate_offsets': 'true',
    'skip': '3',
    'every': '1',
}
result = requests.post('https://enterprise.audd.io/', data=data)
print(result.text)

Go

With the SDK:

npm install @audd/sdk

import { AudD } from "@audd/sdk";

const audd = new AudD("your-api-token");
const matches = await audd.recognizeEnterprise("https://example.com/full-show.mp3", { limit: 10 });
for (const m of matches) console.log(`${m.timecode}  ${m.artist} — ${m.title}`);

Full Node reference →

Or send the request yourself, without the SDK:

var axios = require("axios");

var data = {
    'api_token': 'your api token',
    'url': 'https://audd.tech/djatwork_example.mp3',
    'accurate_offsets': 'true',
    'skip': '3',
    'every': '1',
};

axios({
    method: 'post',
    url: 'https://enterprise.audd.io/',
    data: data,
    headers: { 'Content-Type': 'multipart/form-data' },
})
.then((response) => {
    console.log(response);
})
.catch((error) =>  {
    console.log(error);
});

Rust

With the SDK:

go get github.com/AudDMusic/audd-go

package main

import (
    "fmt"
    "log"

    audd "github.com/AudDMusic/audd-go"
)

func main() {
    client := audd.NewClient("your api token")
    defer client.Close()

    skip := 3
    every := 1
    accurate := true
    matches, err := client.RecognizeEnterprise("https://audd.tech/djatwork_example.mp3", &audd.EnterpriseOptions{
        Skip: &skip,
        Every: &every,
        AccurateOffsets: &accurate,
    })
    if err != nil {
        log.Fatal(err)
    }
    for _, m := range matches {
        fmt.Printf("%s  %s — %s\n", m.Timecode, m.Artist, m.Title)
    }
}

Full Go SDK reference →

PHP

With the SDK:

cargo add audd

use audd::EnterpriseOptions;

let client = audd::Client::builder().api_token("your-api-token").build()?;
let matches = client.recognize_enterprise(
    "https://example.com/full-show.mp3",
    EnterpriseOptions { limit: Some(10), ..Default::default() },
).await?;
for m in matches { println!("{}  {} — {}", m.timecode, m.artist, m.title); }

Full Rust reference →

Swift

With the SDK:

composer require audd/audd

<?php
$audd = new AudD\AudD('your-api-token');
$matches = $audd->recognizeEnterprise('https://example.com/full-show.mp3', ['limit' => 10]);
foreach ($matches as $m) {
    echo "{$m->timecode}  {$m->artist} — {$m->title}\n";
}

Full PHP reference →

Or send the request yourself, without the SDK:

<?php
$data = [
    'api_token' => 'your api token',
    'url' => 'https://audd.tech/djatwork_example.mp3',
    'accurate_offsets' => 'true',
    'skip' => '3',
    'every' => '1',
];
$ch = curl_init();
curl_setopt($ch, CURLOPT_POST, 1);
curl_setopt($ch, CURLOPT_URL, 'https://enterprise.audd.io/');
curl_setopt($ch, CURLOPT_RETURNTRANSFER, true);
curl_setopt($ch, CURLOPT_POSTFIELDS, $data);
$result = curl_exec($ch);
curl_close($ch);
echo $result;

Kotlin

With the SDK:

// Package.swift, in dependencies:
.package(url: "https://github.com/AudDMusic/audd-swift", from: "1.5.19"),

let audd = try AudD(apiToken: "your-api-token")
let matches = try await audd.recognizeEnterprise(
    "https://example.com/full-show.mp3",
    options: .init(limit: 10)
)
for m in matches { print("\(m.timecode)  \(m.artist) — \(m.title)") }

Full Swift reference →

C#

With the SDK:

// build.gradle.kts
implementation("io.audd:audd-kotlin:1.5.16")

val audd = AudD("your-api-token")
val matches = audd.recognizeEnterprise(
    Source.Url("https://example.com/full-show.mp3"),
    limit = 10,
)
for (m in matches) println("${m.timecode}  ${m.artist} — ${m.title}")

Full Kotlin reference →

Java

With the SDK:

dotnet add package AudD

using var audd = new AudD("your-api-token");
var matches = await audd.RecognizeEnterpriseAsync(
    "https://example.com/full-show.mp3",
    new EnterpriseOptions { Limit = 10 });
foreach (var m in matches) Console.WriteLine($"{m.Timecode}  {m.Artist} — {m.Title}");

Full C# reference →

C

With the SDK:

<!-- pom.xml -->
<dependency>
  <groupId>io.audd</groupId>
  <artifactId>audd</artifactId>
  <version>1.5.16</version>
</dependency>

try (AudD audd = new AudD("your-api-token")) {
    var opts = new EnterpriseOptions.Builder().limit(10).build();
    List<EnterpriseMatch> matches = audd.recognizeEnterprise(
        "https://example.com/full-show.mp3", opts);
    for (var m : matches) {
        System.out.println(m.timecode() + "  " + m.artist() + " — " + m.title());
    }
}

Full Java reference →

Or send the request yourself, without the SDK:

// requires OkHttp
OkHttpClient client = new OkHttpClient();
RequestBody data = new MultipartBody.Builder().setType(MultipartBody.FORM)
    .addFormDataPart("api_token", "your api token")
    .addFormDataPart("url", "https://audd.tech/djatwork_example.mp3")
    .addFormDataPart("accurate_offsets", "true")
    .addFormDataPart("skip", "3")
    .addFormDataPart("every", "1").build();
Request request = new Request.Builder().url("https://enterprise.audd.io/")
    .post(data).build();
Response response = client.newCall(request).execute();
System.out.println(response.body().string());

C++

With the SDK:

# CMakeLists.txt
FetchContent_Declare(audd
    GIT_REPOSITORY https://github.com/AudDMusic/audd-c.git
    GIT_TAG        v1.5.16)
FetchContent_MakeAvailable(audd)
target_link_libraries(your_app PRIVATE audd)

audd_client_t* client = audd_client_new("your-api-token", NULL);
audd_enterprise_options_t opts = audd_enterprise_options_default();
opts.limit = 10;
audd_enterprise_match_t* matches = NULL;
size_t n = 0;
audd_recognize_enterprise(client, audd_source_url("https://example.com/full-show.mp3"),
                          &opts, &matches, &n);
for (size_t i = 0; i < n; ++i) {
    printf("%s  %s — %s\n", matches[i].timecode, matches[i].artist, matches[i].title);
}
audd_enterprise_matches_free(matches, n);
audd_client_free(client);

Full C reference →

curl

With the SDK:

# CMakeLists.txt (after installing audd-cpp)
find_package(audd CONFIG REQUIRED)
target_link_libraries(your_app PRIVATE audd::audd)

audd::AudD client("your-api-token");
audd::EnterpriseOptions opts; opts.limit = 10;
auto matches = client.recognize_enterprise(
    audd::Source::url("https://example.com/full-show.mp3"), opts);
for (auto& m : matches) {
    std::cout << m.timecode << "  " << m.artist << " — " << m.title << "\n";
}

Full C++ reference →

Send a local file

Python

Plain HTTP, no SDK:

curl https://enterprise.audd.io/ \
    -F api_token='your api token' \ 
    -F file=@/path/to/largeFile.mp3 \
    -F accurate_offsets='true' \
    -F skip='3' \
    -F every='1'

Node

With the SDK:

pip install audd

from audd import AudD

audd = AudD("your-api-token")
matches = audd.recognize_enterprise("/path/to/full-show.mp3", limit=10)
for m in matches:
    print(f"{m.timecode}  {m.artist} — {m.title}")

Full Python reference →

Or send the request yourself, without the SDK:

import requests
data = {
    'api_token': 'your api token',
    'accurate_offsets': 'true',
    'skip': '3',
    'every': '1',
}
files = {
    'file': open('/path/to/largeFile.mp3', 'rb'),
}
result = requests.post('https://enterprise.audd.io/', data=data, files=files)
print(result.text)

Go

With the SDK:

npm install @audd/sdk

import { AudD } from "@audd/sdk";

const audd = new AudD("your-api-token");
const matches = await audd.recognizeEnterprise("/path/to/full-show.mp3", { limit: 10 });
for (const m of matches) console.log(`${m.timecode}  ${m.artist} — ${m.title}`);

Full Node reference →

Or send the request yourself, without the SDK:

var axios = require("axios");
var fs = require('fs');

var data = {
    'api_token': 'your api token',
    'file': fs.createReadStream('/path/to/largeFile.mp3'),
    'accurate_offsets': 'true',
    'skip': '3',
    'every': '1',
};

axios({
    method: 'post',
    url: 'https://enterprise.audd.io/',
    data: data,
    headers: { 'Content-Type': 'multipart/form-data' },
})
.then((response) => {
    console.log(response);
})
.catch((error) =>  {
    console.log(error);
});

Rust

With the SDK:

go get github.com/AudDMusic/audd-go

package main

import (
    "fmt"
    "log"
    "os"

    audd "github.com/AudDMusic/audd-go"
)

func main() {
    client := audd.NewClient("your api token")
    defer client.Close()

    f, err := os.Open("/path/to/largeFile.mp3")
    if err != nil {
        log.Fatal(err)
    }
    defer f.Close()

    skip := 3
    every := 1
    accurate := true
    matches, err := client.RecognizeEnterprise(f, &audd.EnterpriseOptions{
        Skip: &skip,
        Every: &every,
        AccurateOffsets: &accurate,
    })
    if err != nil {
        log.Fatal(err)
    }
    for _, m := range matches {
        fmt.Printf("%s  %s — %s\n", m.Timecode, m.Artist, m.Title)
    }
}

Full Go SDK reference →

PHP

With the SDK:

cargo add audd

use audd::EnterpriseOptions;

let client = audd::Client::builder().api_token("your-api-token").build()?;
let matches = client.recognize_enterprise(
    std::path::Path::new("/path/to/full-show.mp3"),
    EnterpriseOptions { limit: Some(10), ..Default::default() },
).await?;
for m in matches { println!("{}  {} — {}", m.timecode, m.artist, m.title); }

Full Rust reference →

Swift

With the SDK:

composer require audd/audd

<?php
$audd = new AudD\AudD('your-api-token');
$matches = $audd->recognizeEnterprise('/path/to/full-show.mp3', ['limit' => 10]);
foreach ($matches as $m) {
    echo "{$m->timecode}  {$m->artist} — {$m->title}\n";
}

Full PHP reference →

Or send the request yourself, without the SDK:

<?php
$data = [
    'api_token' => 'your api token',
    'file' => curl_file_create('/path/to/largeFile.mp3', 
        'application/octet-stream', 'file'),
    'accurate_offsets' => 'true',
    'skip' => '3',
    'every' => '1',
];
$ch = curl_init();
curl_setopt($ch, CURLOPT_POST, 1);
curl_setopt($ch, CURLOPT_URL, 'https://enterprise.audd.io/');
curl_setopt($ch, CURLOPT_RETURNTRANSFER, true);
curl_setopt($ch, CURLOPT_POSTFIELDS, $data);
$result = curl_exec($ch);
curl_close($ch);
echo $result;

Kotlin

With the SDK:

// Package.swift, in dependencies:
.package(url: "https://github.com/AudDMusic/audd-swift", from: "1.5.19"),

let audd = try AudD(apiToken: "your-api-token")
let file = URL(fileURLWithPath: "/path/to/full-show.mp3")
let matches = try await audd.recognizeEnterprise(.file(file), limit: 10)
for m in matches { print("\(m.timecode)  \(m.artist) — \(m.title)") }

Full Swift reference →

C#

With the SDK:

// build.gradle.kts
implementation("io.audd:audd-kotlin:1.5.16")

val audd = AudD("your-api-token")
val matches = audd.recognizeEnterprise(
    Source.FilePath(File("/path/to/full-show.mp3")),
    limit = 10,
)
for (m in matches) println("${m.timecode}  ${m.artist} — ${m.title}")

Full Kotlin reference →

Java

With the SDK:

dotnet add package AudD

using var audd = new AudD("your-api-token");
var matches = await audd.RecognizeEnterpriseAsync(
    "/path/to/full-show.mp3",
    new EnterpriseOptions { Limit = 10 });
foreach (var m in matches) Console.WriteLine($"{m.Timecode}  {m.Artist} — {m.Title}");

Full C# reference →

C

With the SDK:

<!-- pom.xml -->
<dependency>
  <groupId>io.audd</groupId>
  <artifactId>audd</artifactId>
  <version>1.5.16</version>
</dependency>

try (AudD audd = new AudD("your-api-token")) {
    var opts = new EnterpriseOptions.Builder().limit(10).build();
    List<EnterpriseMatch> matches = audd.recognizeEnterprise(
        Path.of("/path/to/full-show.mp3"), opts);
    for (var m : matches) {
        System.out.println(m.timecode() + "  " + m.artist() + " — " + m.title());
    }
}

Full Java reference →

Or send the request yourself, without the SDK:

// requires OkHttp
final MediaType MEDIA_TYPE_MP3 = MediaType.get("audio/mpeg; charset=utf-8");
File file = new File("/path/to/largeFile.mp3");
OkHttpClient client = new OkHttpClient();
RequestBody data = new MultipartBody.Builder().setType(MultipartBody.FORM)
    .addFormDataPart("api_token", "your api token")
    .addFormDataPart("file", file.getName(), 
        RequestBody.Companion.create(file, MEDIA_TYPE_MP3))
    .addFormDataPart("accurate_offsets", "true")
    .addFormDataPart("skip", "3")
    .addFormDataPart("every", "1").build();
Request request = new Request.Builder().url("https://enterprise.audd.io/")
    .post(data).build();
Response response = client.newCall(request).execute();
System.out.println(response.body().string());

C++

With the SDK:

# CMakeLists.txt
FetchContent_Declare(audd
    GIT_REPOSITORY https://github.com/AudDMusic/audd-c.git
    GIT_TAG        v1.5.16)
FetchContent_MakeAvailable(audd)
target_link_libraries(your_app PRIVATE audd)

audd_client_t* client = audd_client_new("your-api-token", NULL);
audd_enterprise_options_t opts = audd_enterprise_options_default();
opts.limit = 10;
audd_enterprise_match_t* matches = NULL;
size_t n = 0;
audd_recognize_enterprise(client, audd_source_path("/path/to/full-show.mp3"),
                          &opts, &matches, &n);
for (size_t i = 0; i < n; ++i) {
    printf("%s  %s — %s\n", matches[i].timecode, matches[i].artist, matches[i].title);
}
audd_enterprise_matches_free(matches, n);
audd_client_free(client);

Full C reference →

curl

With the SDK:

# CMakeLists.txt (after installing audd-cpp)
find_package(audd CONFIG REQUIRED)
target_link_libraries(your_app PRIVATE audd::audd)

audd::AudD client("your-api-token");
audd::EnterpriseOptions opts; opts.limit = 10;
auto matches = client.recognize_enterprise(
    audd::Source::path("/path/to/full-show.mp3"), opts);
for (auto& m : matches) {
    std::cout << m.timecode << "  " << m.artist << " — " << m.title << "\n";
}

Full C++ reference →

Fields referring to time

timecode is the position in the original song where the matched fragment is playing (e.g., that can be 00:45 of Imagine Dragons - Warriors).

offset is the position, in the audio file you submitted, of the start of the 12-second-long fragment that contains matches.

start_offset and end_offset are the positions (in milliseconds) in the 12-second-long fragment of the start and end of the fragment we matched to a song.

How to get additional metadata, e.g., ISRCs and UPCs

You don't have to have an enterprise account in order to use the enterprise endpoint. But you need an enterprise account to get ISRCs and UPCs with the results. Let us know if you want access to these features: send an email to api@audd.io.

---

For AI chatbots/agents: Markdown versions of these docs are available.

HTTP API: /.md · /streams.md · /enterprise.md

SDKs: /sdks.md · /sdks/python.md · /sdks/node.md · /sdks/go.md · /sdks/rust.md · /sdks/php.md · /sdks/swift.md · /sdks/kotlin.md · /sdks/dotnet.md · /sdks/java.md · /sdks/c.md · /sdks/cpp.md

Ask about the API: