Speech to text and writing tools

Whisper Dictation types out English speech as you talk, with the Whisper model (Tiny, Base, Small or Large) running in your browser on WebGPU or WebAssembly. The audio and the words never leave the browser, and earlier sentences are transcribed while you are still speaking, so the text is nearly done when you stop.

1. Choose a Whisper model and allow microphone access.
2. Speak in English while the model transcribes earlier sentences.
3. Stop, check the transcript and copy the words into your document.

Larger models need more memory and download time. Use a smaller model on a constrained device and check dictated names and numbers.

Meeting Notes records a meeting, or opens a recording of one, and writes it down speaker by speaker: Whisper transcribes the English, and two speaker models (pyannote segmentation and WeSpeaker voiceprints) tell the voices apart. You name each speaker once at the end and copy or download the notes as text; everything runs in your browser, so the audio and the words never leave it.

1. Record a meeting or open a supported recording.
2. Let Whisper transcribe the English speech and the speaker models separate voices.
3. Name each speaker, check the transcript and export the notes as text.

Overlapping voices and noisy audio can affect both words and speaker labels. Review the output before treating it as meeting minutes.

Essay Repetition Checker finds the words and word families an essay leans on, used often or close together, ignoring common words like "the", and highlights every use so you can edit it in place. The essay is checked in your browser and never leaves the page.

1. Paste your draft into the editor.
2. Review frequent or nearby repeated words and word families.
3. Edit in context, keeping deliberate repetition and revising the uses that weaken the writing.

Frequent words are a prompt for review, not automatic errors. Subject-specific terms may need to repeat for clarity.