Frequently Asked Questions (FAQ)

General

What is Txtify?

Txtify is a FREE and OPEN-SOURCE tool that converts audio and video into text using state-of-the-art AI models for rapid and precise transcriptions. With Txtify, you can convert your files or urls effortlessly, and it's available for self-hosting, offering you full control over your transcription process.


Models

Which AI models are supported?

Txtify utilizes advanced AI models from Whisper and SeamlessM4T, including Whisper Tiny, Whisper Base, Whisper Small, Whisper Medium, Whisper Large, SeamlessM4T Medium, and SeamlessM4T Large. These models are sourced from Hugging Face repository.


What are the memory requirements and speed for the models?

Below is a table summarizing the available Whisper models, their memory requirements, and their relative inference speeds. The speed and memory requirements may vary based on the available hardware.


Size Parameters Multilingual model Required VRAM Relative speed
tiny 39 M tiny ~1 GB ~32x
base 74 M base ~1 GB ~16x
small 244 M small ~2 GB ~6x
medium 769 M medium ~5 GB ~2x
large 1550 M large ~10 GB ~1x

Languages

Which languages are supported for transcription?

Txtify supports transcription in the following languages: Afrikaans, Amharic, Arabic, Assamese, Azerbaijani, Belarusian, Bulgarian, Bengali, Bosnian, Catalan, Cebuano, Czech, Welsh, Danish, German, Greek, English, Spanish, Estonian, Persian, Finnish, French, Galician, Gujarati, Hebrew, Hindi, Croatian, Hungarian, Armenian, Indonesian, Icelandic, Italian, Japanese, Javanese, Georgian, Kazakh, Khmer, Kannada, Korean, Lao, Lithuanian, Latvian, Malayalam, Mongolian, Marathi, Malay, Burmese, Nepali, Dutch, Punjabi, Polish, Portuguese, Romanian, Russian, Sinhala, Slovak, Slovenian, Albanian, Serbian, Swedish, Swahili, Tamil, Telugu, Thai, Turkish, Ukrainian, Urdu, Uzbek, Vietnamese, Yiddish, Yoruba, Chinese.


Which languages are supported for translation?

Translations are supported in the following languages using DeepL: Arabic, Bulgarian, Czech, Danish, German, Greek, English, English (British), English (American), Spanish, Estonian, Finnish, French, Hungarian, Indonesian, Italian, Japanese, Korean, Lithuanian, Latvian, Norwegian Bokmål, Dutch, Polish, Portuguese, Portuguese (Brazilian), Portuguese (excluding Brazilian Portuguese), Romanian, Russian, Slovak, Slovenian, Swedish, Turkish, Ukrainian, Chinese (simplified).


Limitations

Are there any limitations?

Yes, this version has limitations. You can upload files up to 100MB or transcribe YouTube videos up to 10 minutes. However, when you self-host Txtify, you can modify and run the application without these limitations, giving you full control over the transcription process.


File and Process Deletion

What happens to my files and processes after I close the window?

After the window is closed, all generated files and the transcription process are automatically deleted to ensure your data privacy and security.


Self-Hosting

Can I self-host Txtify?

Yes, you can self-host Txtify with full features. Please check the GitHub repo for instructions.


Demo Video

Is there a demo video showcasing the features of Txtify?

Yes, you can watch the demo video below:


Contribute

Want to contribute to Txtify?

Your contributions are welcome! Feel free to open a pull request on our GitHub repository.


Report Errors

Found any issues or bugs?

Please report them using the contact form on the contact page. We appreciate your feedback and will work to resolve any problems as quickly as possible.


Contact

If your question wasn't answered here, please use our contact page to reach out to us.