Go back
Asynchronous Speech-To-Text API Documentation
v1
Added:
13/01/2021
Updated at:
02/08/2021
Rev.ai provides quality speech-text recognition via a RESTful API. All public methods and objects are documented here for developer reference.
For a real-time speech to text solution, use Rev.ai's [Streaming API](/docs/streaming).
# Base Endpoint
The base url for this version of the API is
> `https://api.rev.ai/speechtotext/v1`
All endpoints described in this documentation are relative to this base url.
# Quick Start
Follow the [getting started checklist](https://www.rev.ai/getting_started)
## Get your Access Token
You can generate your [access token](#section/Authentication/Access-Token) on the [settings page](https://www.rev.ai/access_token) of your account. This access token only needs to be generated once and never expires. You can re-generate your token, however this will invalidate the previous token.
## Submit a File
To submit an audio file for transcription to Rev.ai:
```
curl -X POST "https://api.rev.ai/speechtotext/v1/jobs" -H "Authorization: Bearer $REV_ACCESS_TOKEN" -H "Content-Type: application/json" -d "{\"media_url\":\"https://www.rev.ai/FTC_Sample_1.mp3\",\"metadata\":\"This is a sample submit jobs option\"}"
```
You’ll receive a response like this:
~~~
{
"id": "Umx5c6F7pH7r",
"created_on": "2018-09-15T05:14:38.13",
"name": "sample.mp3",
"metadata": "This is a sample submit jobs option for multipart",
"status": "in_progress"
}
~~~
The `id` (in this case `Umx5c6F7pH7r`) will allow you to retrieve your transcript.
## Get Your Transcript
Once a transcription job's `status` becomes `transcribed`, you can retrieve the transcript in JSON format by running:
```
curl -X GET "https://api.rev.ai/speechtotext/v1/jobs/{id}/transcript" -H "Authorization: Bearer $REV_ACCESS_TOKEN" -H "Accept: application/vnd.rev.transcript.v1.0+json"
```
Alternatively you can get the plain text version by running:
```
curl -X GET "https://api.rev.ai/speechtotext/v1/jobs/{id}/transcript" -H "Authorization: Bearer $REV_ACCESS_TOKEN" -H "Accept: text/plain"
```
You can poll for the `status` of your job by querying for the job periodically:
```
curl -X GET https://api.rev.ai/speechtotext/v1/jobs/{id} -H "Authorization: Bearer $REV_ACCESS_TOKEN"
```
**Note:** Polling is NOT recommended in a production server. Rather, use [webhooks](#section/Webhooks) to asynchronously receive notifications once the transcription job completes
If you have any further questions, contact us at
# Submitting Files
Two `POST` request formats can be used to submit a file: `application/json` or `multipart/form-data`.
## JSON
This is the preferred method of file submission. Uses the `media_url` property to provide a direct download URL to the Rev.ai server.
This method supports the use of pre-signed URLs. Links to videos hosted on platforms like Youtube are not valid because they are not direct download links.
**Important note on presigned URLs:**
Signed URLs usually have an expiration time which is configurable. To ensure the Rev.ai server can access the link, make sure the expiration time is set to 2 hours or more.
In the event you plan on resending the same file, make sure to generate a new presigned URL.
## FormData
Used to send a local file to the Rev.ai server. This allows the customer to send the file directly from the host machine.
Certain limits apply to this format, see the [Async API Limits section](#section/Async-API-Limits) for more detals.
# Turnaround Time and Chunking
Often, especially for shorter files, your transcript will be ready in 5 minutes or less. It generally takes no longer than 15 minutes to return longer audios. If you require faster turn around time please contact
Chunking is the act of breaking audio files into smaller segments. Rev.ai uses this method to decrease turnaround time of audios greater than 3 minutes in length.
# Webhooks
If the optional `callback_url` is provided, the API will make an HTTP POST request to the `callback_url` with the
following request body when the job either completes successfully or fails.
## Sample Webhook
**On Successful Transcription Job**
```
{
"job": {
"id": "Umx5c6F7pH7r",
"status": "transcribed",
"created_on": "2018-05-05T23:23:22.29Z",
"callback_url": "https://www.example.com/callback",
"duration_seconds": 356.24,
"media_url": "https://www.rev.ai/FTC_Sample_1.mp3"
}
}
```
**On Failed Transcription Job**
```
{
"job": {
"id": "Umx5c6F7pH7r",
"status": "failed",
"created_on": "2018-05-05T23:23:22.29Z",
"callback_url": "https://www.example.com/callback",
"failure": "download_failure",
"failure_detail": "Failed to download media file. Please check your url and file type"
}
}
```
**Important notes for using webhooks:**
The API will make a POST request, not a GET request, to the `callback_url`. The request body is the job details.
You can unsubscribe from a webhook by responding to the webhook request with a 200 response.
If a webhook invocation does not receive a 200 Rev.ai will retry the `callback_url` every 30 minutes until either 24 hours have passed or we receive a 200 response.
For initial webhook testing, you can try using a third party webhook testing tool such as [https://webhook.site/](https://webhook.site/).
# Async API Limits
The following default limits apply per user, per endpoint and are configurable by Rev.ai support. If you have any further questions, contact us at
- 10,000 transcription requests submitted every 10 minutes
- 500 transcriptions processed every 10 minutes
- Multi-part/form-data requests to the /jobs endpoint have a concurrency limit of 10 and a file size limit of 2GB
- POST requests to the /jobs endpoint that use the media_url property do not have a concurrency limit or file restriction. They are only limited by the first two bullet points
# Error Handling
The API indicates failure with 4xx and 5xx HTTP status codes. 4xx status codes indicate an error due to the request provided (e.g., a required parameter was omitted). 5xx error indicate an error with Rev.ai's servers.
When an 4xx error occurs during invocation of a request, the API responds with a [problem details](https://tools.ietf.org/html/rfc7807) as HTTP response payload.
The problem details information is represented as a JSON object with the following optional properties:
| Property | Description |
|------------|-----------------------------------------------|
| type | a URI representing the type for the error |
| title | a short human readable description of type |
| details | additional details of the error |
| status | HTTP status code of the error |
In addition to the properties listed above, the problem details object may list additional properties that help to troubleshoot the problem.
## Example Errors
```
// Bad Submit Job Request
{
"parameter": {
"media_url": [
"The media_url field is required"
]
},
"type": "https://www.rev.ai/api/v1/errors/invalid-parameters",
"title": "Your request parameters didn't validate",
"status": 400
}
// Invalid Transcript State
{
"allowed_values": [
"transcribed"
],
"current_value": "in_progress",
"type": "https://rev.ai/api/v1/errors/invalid-job-state",
"title": "Job is in invalid state",
"detail": "Job is in invalid state to obtain the transcript",
"status": 409
}
```
## Retrying Failed Requests
Some errors can be resolved simply by retrying the request. The following error codes are likely to be resolved with successive retries.
| Status Code | Error |
|---|:---|
| 429 | Too Many Requests |
| 502 | Bad Gateway |
| 503 | Service Unavailable |
| 504 | Gateway Timeout |
Note: With the exception of the 429 status code, it is recommended that the maximum number of retries be limited to 5 attempts per request. The number of retries can be higher
for 429 errors but if you notice consistent throttling please contact us at .
