Whisper Web
Whisper Web accepts audio and video uploads, microphone recordings and public YouTube links, then returns a transcript with timestamps and speaker labels. It also produces summaries with key points and action items. Whisper Web is an independent product and says it is not affiliated with or endorsed by OpenAI.
Upload a recording and keep the timestamps
The service lists MP3, MP4, M4A, WAV and other recording formats. Its references to Zoom, Teams and Google Meet concern exported recordings; you should not assume a meeting bot or live capture integration from that description. Obtain permission to record and upload a conversation before sending it.
TXT and DOCX exports suit editing, while SRT and VTT exports suit subtitle workflows. Speaker labels help you navigate a conversation, but names, quotations and decisions still need a check against the recording. A summary can omit qualifications or assign an action to the wrong person. Keep the timestamped transcript available while editing the shorter version.
The vendor advertises more than 100 languages and 98% accuracy on clear audio. Those are product claims, not results from an independent test supplied here. Try a representative recording with your accents, overlap and background noise before processing a large archive.
Check the free allowance against pricing
The homepage's broad free messaging and 2GB upload description differ from the more specific pricing page. Pricing lists two free uploads, a 500MB free-file limit and the first ten minutes free per file, plus three complimentary AI summaries. Use those narrower published limits when planning an initial trial, and confirm what the upload screen allows.
Pro is listed at $12.99 per month with 1,200 transcription minutes and a 2GB file limit. Each minute of audio consumes one minute of quota. AI summaries are unlimited and do not count toward the quota. Unlimited summaries therefore do not mean unlimited audio processing. The page also lists twelve specialized templates and priority processing for Pro.
Review privacy and refund terms
Whisper Web says it encrypts uploads, deletes files after processing and does not train models on customer audio. Those statements are vendor assurances rather than an independent security audit. Browser-based access does not establish that the transcription runs only on your device. Check the service's current terms before uploading confidential interviews or regulated information.
The published refund window is fourteen days for an initial paid subscription, with used audio deducted at $0.035 per minute. That deduction differs from the promotional per-minute figure displayed for Pro. Save the current terms alongside a purchase, and test exports and speaker labeling on a short, non-sensitive recording first.