Audio captcha solver: How to bypass audio reCAPTCHA and any other audio captcha
2Captcha launched an audio recognition service. The audio method can be used to bypass audio captchas or to recognize any audio record. The service uses an automatic audio recognition system.
Motivation: Audio captchas typically restrict access to services and information based on our social and cultural backgrounds, physical and cognitive abilities, the devices or networks we connect from, and other criteria. An audio recording can be converted to text using auditory recognition technology. Audio recognition method allows to convert an audio record to text.
The limitations are:
- Maximum file size: 1 megabyte
- Audio duration: not limited
- Supported audio format: mp3 only
- Supported speech languages: English, French, German, Greek, Russian, Portuguese
The recognition is fully automated and performed by a neural network trained for speech recognition.
How to recognize audio to text
To recognize audio in text, you need to execute 2 queries:
- Submit a file for recognition;
- Get a result.
Submit a file
Collect the audio captcha parameters.
Audio recognition method allows bypassing the audio captcha challenge automatically.
To successfully bypass the captcha user need to send values:
body
- base64 encoded audio file in mp3 formatlang
- the language of audio record
Interact with audio captcha bypass API
When you have all required values submit the captcha to in.php
endpoint of API with the following parameters:
parameter | description |
---|---|
key | API key |
method | audio - indicates that sending audio |
body | base64 encoded audio file in mp3 format |
lang | The language of audio record. Supported languages are: 'en' : English 'fr' : French'de' : German'el' : Greek'pt' : Portuguese'ru' : Russian |
Request body example:
{
"key":"xxxxx",
"method":"audio",
"body":"",
"lang":"pt",
"json":1
}
If everything is OK you will receive the response with your request ID.
Example:
{
"status":1,
"request":"xxxxx"
}
If the file upload failed, an error response should be returned. A list of all errors with a description is collected here.
To recognize an audio user needs to encode the mp3 file to base64.
Make a 15-20 seconds timeout then submit a HTTP GET request to API URL res.php
to get the result.
Get a result
Use the returned id of audio captcha to make request to res.php
API endpoint with the following parameters:
col | description |
---|---|
key | API key |
action | get - get the answer for captcha |
id | ID of captcha returned by in.php |
Example:
{
"key":"xxxxx",
"action":"get",
"id":5
}
If audio captcha is already recognized server will return the text in the following format:
{
"status": 1,
"request": "hello world"
}
If the recognition process is not finished yet, the server will return CAPCHA_NOT_READY
result.
If something goes wrong, the server will return an error code.
A list of all errors with a description is collected here.
Pricing
The price for 1000 recognitions is $0.5.
Helpful links
Audio captcha recognition code examples: