Text to Speech

Yes. You can generate and download speech for free, with no subscription, account or usage-count limit.

Yes. You can use the generated audio in videos, podcasts, courses, advertisements and other commercial projects without paying a commercial license fee. Use text you own or have permission to use.

Choose the language of your text, then select one of its available voices. Voices are designed for particular languages and accents; pronunciation and quality can vary. Numbers, abbreviations and mixed-language text may need review.

AI voices generate audio on your device without uploading text to our servers. System voices may be local or online; online voices may send text to their speech provider.

No installation is needed. AI voices download the required speech resources on first use, so preparation may take longer. Later uses are usually quicker.

There is no fixed text-length limit. Longer text needs more processing time and memory and may fail on devices with limited resources. If generation fails, split the text into smaller sections or close other tabs.

Open Settings and start with speed at 1×, pitch at 0 and volume at 100%. Lower the speed for slower narration or raise it for a quicker pace. Pitch makes the voice lower or higher without changing speed. Make small changes and preview them; large adjustments can sound less natural, and high volume can distort the sound.

Choose Normal for continuous narration, voiceovers, or simply listening to and downloading speech. Read-along divides the text into shorter phrases and highlights the current passage in the input area during playback, making it useful for following a text, language practice, or checking pronunciation. After generation, the text is read-only: click a phrase to jump to its audio, or select Edit text to make changes. Highlighting is phrase-level, not word-by-word, and pauses may differ between modes. Both modes support audio downloads.

Choose Read-along mode to download SRT or VTT after generating speech. These subtitles use phrase-level timing adjusted to the exported audio and speed; they are not word-by-word captions. Edit subtitles opens the current audio and subtitles in Subtitle Editor for review. Normal mode does not provide timed subtitles.

Yes. Import TXT, MD, DOCX, PDF, EPUB or ODT. The extracted text appears in the editable input. PDF lets you choose pages, and EPUB lets you choose chapters. Documents are processed in your browser. Review reading order, tables and special characters before generating speech. MD imports its original text.

Yes. Import the PDF, choose the pages and scan language, then select Recognize scanned text. Text recognition runs only when you request it, and processes pages without selectable text. Review the recognized text, especially names, numbers and columns, before generating speech.

Files up to 100 MB can be imported, with a warning above 20 MB. These limits do not guarantee that every device can process a large file or generate long speech. Use page or chapter selection, split long texts, and close other tabs if processing fails.

AI voices generate downloadable audio in MP3, WAV, M4A or FLAC, with optional read-along and subtitles. System voices use the speech voices available on your device or browser and support playback, pause and stop, without an audio download. Available languages and voices vary by device.

Audio and text are saved in this browser without automatic expiry. Clearing browser data or running out of space may remove them. Download important recordings.