how it works
Free voice typing in your browser. Upload audio and video for AI transcription. Extract YouTube transcripts, generate speech, automate with the API or MCP, and save work across devices.
// quick start — 3 steps
pick your language
Select from 55+ languages. Hindi, Spanish, Arabic, French, German, and more — all free.
click start recording
Allow microphone access once. Speech recognition runs via your browser's built-in engine — not uploaded to SHRP.
speak, edit and save
Watch words appear in real-time. Edit, copy, download, or save to My Projects for later.
// microphone mode (free)
open shrp in your browser
No download, no install. Works on Chrome, Edge, Safari, and Firefox. No account required.
allow microphone access
Your browser asks for permission once. Click Allow. You can revoke it any time from browser settings.
speak naturally
Words appear in real time as you speak. Uses your browser's built-in speech engine — not uploaded to SHRP.
copy, download or save
Copy to clipboard, download as .txt, or save to My Projects. Auto-saved to browser — nothing lost if you close the tab.
use ai enhance
Click AI Enhance to fix grammar, add paragraphs, or both. Free for everyone, no account needed. Powered by Claude Haiku.
// file upload mode (starter+)
upload your audio or video file
Drag and drop or click to browse. Supports MP3, WAV, M4A, MP4, MOV, OGG, FLAC, WebM — up to 500MB on Pro.
assemblyai processes it
Your file is sent securely to AssemblyAI — one of the most accurate transcription engines available, delivering ~98% accuracy on clean audio.
speaker diarization
If your audio has multiple speakers, SHRP automatically detects and labels each one. Speaker labels appear in the transcript and all exports.
read the confidence heatmap
Words are colour-coded by AI confidence. Yellow = double-check, Red = verify before publishing. Instantly see where to review.
use ai extraction
One click to extract meeting minutes, action items, key names and dates, or a full summary from your transcript.
export in any format
Download as TXT, SRT, VTT, DOCX, or JSON. SRT and VTT include timestamps for video captioning. JSON includes full word-level data.
// youtube transcripts
paste a youtube link
Open YouTube to Text and paste the video URL. SHRP checks whether an available transcript can be retrieved for the video.
extract readable text
When a transcript is available, SHRP turns it into clean text you can copy, save, or reuse.
send it into your workflow
Use the transcript for notes, summaries, research, blog drafts, or automation through the API. You can also click Generate suggested actions or ask a custom transcript prompt.
save generated outputs
Running a quick action, suggested action, or custom prompt counts as a smart transcript action. Signed-in outputs are saved to history so you can reopen them later.
// text to speech
go to text to speech
Navigate to /text-to-speech. No login required for basic use.
type or paste your text
Enter any text — articles, scripts, emails, study notes, anything.
choose your voice and language
Select from 60+ languages and 2,000+ voices — preview and hear each one before you generate. Includes Neural2, Studio, Journey, and Chirp3 HD.
generate and download
Click Generate. Audio plays instantly in your browser. Download as MP3 for offline use.
// smart voice tools
dictate naturally and auto-format into a structured document. no editing needed.
meeting notes
Structured notes with key points and decisions.
action items
Extract tasks and next steps from any discussion.
email draft
Turn a voice memo into a professional email.
summary
Condense long transcripts into a short summary.
soap notes
Medical SOAP note format for clinical documentation.
rewrite
Clean up and reformat raw dictation.
// api and mcp
developers and agent users can call SHRP from outside the browser. casual users can ignore this section — it does not change the main speech-to-text workflow.
// my projects — save and sync
every transcription can be saved to My Projects — organised, tagged, and searchable.
no account
Saved in browser only. 25 projects max.
free account
25 cloud projects. Access from any device.
free signupcloud+ ($3/mo)
Unlimited cloud projects. Auto-sync.
from $3/mo// supported file formats
| format | quality | notes |
|---|---|---|
| mp3 | good | most common, widely compatible |
| wav | best | uncompressed, highest accuracy |
| m4a | good | default on iOS and macOS |
| mp4 | good | video with audio track |
| mov | good | Apple QuickTime video |
| ogg | good | open source format |
| flac | best | lossless compression |
| webm | good | browser recordings |
// browser support
desktop
mobile and tablet
// privacy
microphone audio not uploaded to SHRP — Microphone audio is not uploaded to or stored by SHRP — processed via your browser's speech engine
no cookies, no session recording — Cookieless Vercel Analytics only. No GA, no Facebook Pixel, no Clarity
no login required — Start immediately, completely anonymous
uploaded files deleted within 30 days — File transcription uploads are automatically purged from storage
transcripts are yours — We never use your content for advertising or AI training
// tips for best results
use a quiet environment — Background noise reduces accuracy significantly
speak clearly and naturally — Don't rush — natural pace gives better results
use chrome or edge — They have the best Web Speech API implementations
enable auto punctuation — Saves you from manual editing after dictating
use ai enhance after dictating — Fix grammar and formatting in one click — free, no account needed
select the right language — Matching the language improves accuracy considerably
ready to start?
free forever for voice typing. upgrade when you need file transcription or cloud sync.