--- title: "Free Browser AI Tools: No Upload, No Signup | Yaps" description: "Free voice typing, transcription, subtitles and text to speech that run entirely in your browser. No accounts, no uploads, no length limits. Built by the team behind Yaps." canonical: "https://www.yaps.ai/tools" language: "en" --- Free tools No signup · Nothing uploaded # Free AI tools that run in your browser. Real speech models, running inside your browser. Nothing is uploaded, nothing is capped, and there is no account to make. 01 Browser tools ## Browser tools [Live ### Speech to Text Free speech to text in your browser. Tap the mic, talk, and your words appear. No signup, no time limit, and nothing is uploaded. Open tool](/tools/speech-to-text) [Live ### Microphone Test Check your mic with a live waveform, frequency view, device details and a practical quality rating. No audio is uploaded. Open tool](/tools/microphone-test) [Live ### Subtitle Generator Drop in a video and get a clean SRT or VTT back. Cues are timed to the words, editable, and generated without uploading the file. Open tool](/tools/subtitle-generator) [Live ### Audio to Text Convert audio to text free, however long the recording. No length cap, because the model runs in your browser rather than on our servers. Open tool](/tools/audio-to-text) [Live ### Text to Speech Free online text to speech in 31 languages, on the same engine the Yaps app uses. No signup, no watermark, and your text never leaves the page. Open tool](/tools/text-to-speech) [New ### Ebook to Audio Turn an EPUB or PDF into natural-sounding chapter audio. Listen in your browser first, then download the audiobook when it is ready. Open tool](/tools/ebook-to-audio) [New ### Ebook Text to Speech Read an EPUB or PDF aloud with voice, language and chapter controls. Your book stays on your device while Yaps speaks. Open tool](/tools/ebook-text-to-speech) [New ### Ebook to MP3 Turn EPUB or PDF chapters into downloadable audio, with clear progress while each section is synthesized. Chapters save as WAV, ready to convert to MP3. Open tool](/tools/ebook-to-mp3) [New ### PDF to Text Extract readable plain text from a text-based PDF in your browser. Nothing is uploaded, and scanned pages are clearly flagged for OCR. Open tool](/tools/pdf-to-text) [New ### PDF to Markdown Turn PDF headings and paragraphs into portable Markdown for notes, research and writing. Runs locally with no account or upload. Open tool](/tools/pdf-to-markdown) [New ### PDF to Word Export the readable text from a PDF as a .docx Word document. Text comes across cleanly; complex layouts and scans may need review. Open tool](/tools/pdf-to-word) [New ### PDF to DOC Convert a text-based PDF into a modern editable .docx file. Yaps does not generate the obsolete binary .doc format, and nothing is uploaded. Open tool](/tools/pdf-to-doc) [New ### PDF to RTF Extract PDF text into a broadly compatible rich text file for WordPad, TextEdit and other word processors. Runs locally in your browser. Open tool](/tools/pdf-to-rtf) [Live ### Filler Word Counter Paste any transcript and see every um, uh, like, and you know highlighted. Get a speech grade in one glance. Runs entirely in your browser. Open tool](/tools/filler-word-counter) [Live ### Headshot Background Remover Remove the background from a headshot and download a transparent PNG. Built for photos of people, and the photo never leaves your browser. Open tool](/tools/headshot-background-remover) [Live ### Speaking Speed Test A speaking speed test that measures your words per minute both ways. Most people speak about three times faster than they type. Open tool](/tools/speaking-speed-test) ## About these tools ## Free AI tools that run in your browser, not on our servers Every tool on this page does its work on your own machine. The speech tools download a speech recognition model once, the voices and the cutout model work the same way, and from then on the file you are working with never leaves the tab you are in. That is the opposite of how most free tools in these categories work. Normally you upload a recording or a photo, a server processes it, and you get a result back. That model is why those sites need accounts, why they cap your minutes, and why their privacy policy matters to you. Ours have a different shape, and the practical effects follow from it. There is no account, because there is nothing to meter. There are no length limits on transcription, because a longer file costs us nothing. And we cannot read, store or lose your work, because it was never sent to us. ## What you can do here without signing up Dictate. The speech to text tool turns your voice into text as you talk, the microphone test checks the signal before you start, and the speaking speed test measures how much faster you talk than you type. Transcribe. Audio to text takes an existing recording of any length and gives you the words back, with timestamps if you want them. The subtitle generator does the same job but produces a timed .srt or .vtt file for a video. Read and listen. Text to speech reads anything you paste in, in 31 languages, and hands you a WAV file. Ebook to audio turns an EPUB or PDF into chaptered narration, while ebook text to speech lets you read a book aloud in the browser with playback controls. Export. Ebook to MP3 creates downloadable audio chapters from the same local book. The filler word counter finds every um, cliché and wordy phrase in a transcript or a draft. Cut out. The headshot background remover takes a photo of a person and gives you back a transparent PNG. ## Turn an ebook into audio without uploading it Ebook to audio is the quickest way to listen to a book you already own. Drop in an EPUB or PDF, let the page identify its chapters, and start listening as the first section is generated. The text is processed in your browser, so the book itself is not sent to a conversion server. Use ebook text to speech when you want a reader rather than a finished file: pause, change speed, move between chapters and follow along with the text. Use ebook to MP3 when you want audio files you can keep for a flight, a commute or another player. These are three entry points into the same private reading workflow, not separate upload funnels. The first version is designed for text-based, DRM-free books. Scanned pages, DRM-protected files and complicated PDF reading order may need OCR or extra cleanup, so the tool tells you what it can reliably extract before it starts speaking. ## Convert a PDF to text, Markdown, Word or RTF PDF to text is the simplest way to get the words out of a text-based PDF. PDF to Markdown keeps headings and paragraphs useful for notes and research, while PDF to Word and PDF to DOC produce a modern .docx file you can open and continue editing in Word or another office suite. The download is DOCX, not the obsolete binary .doc format. These converters use the same browser-local document pipeline as the read-aloud tools. Your PDF stays on your device while its text is extracted, so there is no upload, account or conversion queue. RTF is a broadly compatible rich-text export for word processors that do not need a full Word package; HTML is useful when you want a standalone document that opens in any browser. Text-based PDFs work best. A scanned PDF is made of page images and needs OCR, while multi-column layouts, tables, footnotes and unusual reading order may require a quick review. PDF to Word preserves readable content rather than promising pixel-perfect visual layout; the source PDF remains the right file for archival fidelity. ## Why an on-device tool can be free without a catch Free tools usually have a business model hiding behind them. Either you are the product, or the free tier is a funnel with a wall placed where it hurts most: after you have uploaded the file, or at minute ten of a fifteen minute recording. Running the model in your browser removes the cost that those walls exist to recover. We are not paying for GPU time, so there is no reason to ration it. The honest version of the catch is this: we build a desktop app, these tools are how people find out we exist, and each one tells you plainly what it cannot do and what the app does instead. The trade you make is a one-off download. The speech model is about 80MB and the voices about 145MB, fetched once and then cached by your browser. After that they start instantly and keep working with no connection at all. ## Do these tools work offline? Once a tool's model is cached, yes. The recognition, the synthesis and the cutout were never using the network, so a lost connection does not stop them. You do need the page itself to load, so open it before you go offline. The filler word counter is the exception that needs nothing at all. It is a dictionary and a set of rules rather than a model, so it works from the moment the page loads. Clearing your browser storage or opening a private window empties the cache, so the model downloads again next time. Different browsers keep separate caches too. ## Which tool do I need? If you want to talk and get text, use speech to text. If you already have a recording, use audio to text. The difference is only whether the audio is live or a file. If the words need to sit on top of a video with timings, you want the subtitle generator rather than audio to text, because it produces a subtitle file with cues timed to the words. If you have text and want audio, that is text to speech, which is the reverse of the first two and a common thing to mix up. If you have a transcript and want it tidier, use the filler word counter. If you want a clean cutout of a person, use the background remover. ## Questions **Are these tools really free?** Yes, with no account and no trial. They run on your own machine, so there is no per-use cost for us to recover. **Is anything I upload stored?** Nothing is uploaded in the first place. Your audio, video, text and photos are read by the page and processed in the tab in front of you. There is no processing server involved, which you can check by turning your wifi off once a tool's model has downloaded. **Why do some tools need a download first?** The speech, voice and cutout tools use real machine learning models, and the model has to reach your machine before it can run. It is a one-off: your browser caches it, so later visits start immediately and work offline. **Is there a file size or length limit?** No. Because the work happens on your hardware rather than ours, there is nothing for us to cap. A two hour recording is fine, it simply takes longer to process than a two minute one. **Do they work on a phone?** They run in mobile browsers, but the model download and the processing are noticeably slower than on a laptop. On Android, the Yaps keyboard is a much better way to dictate than a browser tab. **What is the difference between speech to text and text to speech?** Speech to text listens and writes: you talk, it types. Text to speech reads aloud: you type, it speaks. They are opposite jobs and we have a separate tool for each. **Can I turn an EPUB or PDF into an audiobook?** Yes. Ebook to audio reads a text-based EPUB or PDF chapter by chapter in your browser, and ebook to MP3 lets you download the generated chapters. DRM-protected and scanned books may not be supported yet. **What is the difference between ebook to audio and ebook text to speech?** Ebook text to speech is the reader experience: listen immediately, pause, change speed and follow the text. Ebook to audio is the broader workflow for generating chapter audio, while ebook to MP3 focuses on downloadable files. **Can I convert a PDF to TXT, Markdown or Word?** Yes, for text-based PDFs. PDF to text extracts plain text, PDF to Markdown keeps a useful heading and paragraph structure, and PDF to Word or PDF to DOC exports a modern .docx document. Scanned PDFs need OCR, and complex columns or tables may need a quick review. **What is PDF to RTF useful for?** RTF is a portable rich-text format that opens in many word processors, including WordPad and TextEdit. Yaps generates it locally from selectable PDF text; scanned pages and complex layouts still need OCR or a review after extraction. Start yapping Mac · Windows · Linux · Android ## Or skip the tool and *just talk*. Yaps turns your voice into clean text, live, as you speak. No cleanup pass required. [Download for Mac](https://github.com/richawo/yaps-releases/releases/latest) Requires macOS 13.0+ (Apple Silicon recommended)iOS coming soon