Speech To Text OpenAI request
Speech To Text OpenAI request generates text from audio and sends this text to the bot.
Set up the component
1. From the Components section on the right panel, drag or double-click the Speech To Text OpenAI request component to the desired screen.
2. In the Access token, add the OpenAI token. You can get it in your OpenAI personal account (LogIn API). Then, in the Overview section, open the Personal tab and select the View API keys section in it.
If OpenAl is not available in your country, use a VPN service.
Generate a token using the Create new secret key button.
3. Upload the audio for decryption to the Audio File field. The audio file can be downloaded from a computer or added as a link. MP3 format is supported for audio.
The shorter the audio, the faster the text is generated.
4. Select the language of the audio recording. If you select a language that does not match the audio language, then your audio file may be translated into the selected language in some cases.
5. Write Prompt. This field allows you to improve the quality of audio file recognition in some cases. More information about this parameter can be found in the official OpenAI help →
This field is optional.
6. Add Variable name for the response so that the text transcript of the audio gets into it.
After the text is generated and gets into the bot, it can be output using this variable.
7. Specify the Тhe screen that will be executed if the request succeeds. This screen will be executed upon successful request to the neural network. If the successful request screen is not selected, then if successful, the following component will be executed.
8. Specify the The screen that will be executed if the request fails. This screen will be executed in case of one or more request errors. Examples of errors:
- the text is being generated for too long;
- invalid Open AI token is specified;
- OpenAI is not available;
- the error is on the OpenAI side.
