# WhisperWeb Review Free, browser-local audio and video transcription powered by OpenAI's Whisper model, with no signup or file uploads required. ## Editor's Score: 68/100 Value scores the maximum 25 because WhisperWeb is genuinely free with no paid tier, no watermarks, and no signup, which is rare even among free tools. Ease of use scores 21 because the no-install, drop-a-file workflow is about as frictionless as browser tools get, though the developer's own caveat about unverified mobile support and the need to test short clips first adds real friction for non-desktop users. Capability scores 17 because the hard 20-minute file cap is an architectural ceiling rather than a freemium nudge, the tool is limited to tiny/base/small Whisper models with no access to larger ones, and there is no speaker diarization; Tremigos handles longer files and adds translation in a single workflow. Delivery scores 5 because the developer explicitly discloses that mobile transcription is unverified, that larger files can fail or produce poor results depending on device memory, and that the tool has only been fully tested in one specific browser-OS combination, which is an unusually narrow verified scope for a tool presented as broadly available. ## Pros - Genuinely free with no hidden tiers, watermarks, or signup walls: the 200 MB / 20-minute limit is the only constraint on the free experience. - Audio never leaves the device, which is a real, architecture-level privacy guarantee rather than a policy promise, because the Whisper model runs entirely in the browser. - Subtitle workflow goes beyond raw transcription: SRT and VTT export, format conversion between the two, and subtitle-to-text stripping are all included at no cost. ## Cons - The 20-minute file cap is a hard architectural limit tied to the browser-local model, not a freemium nudge, so there is no paid upgrade path on whisperweb.site to lift it. - Mobile transcription is explicitly unverified by the developer: the site warns that a responsive layout does not mean real-phone recognition has been tested, which is an unusual and honest admission that still leaves mobile users in the dark. - Model accuracy depends on the user's device hardware: a low-memory machine running the small model on a long, accented recording can produce noticeably worse output than a cloud-backed service using Whisper Large, and there is no way to access a larger model through this tool. ## Pricing Free. - Free: $0 Pricing reflects what we saw at the time of review (2026-10). Confirm current pricing on the official site. Category: AI Voice, Audio & Music (https://aitoolseekers.com/category/ai-voice-audio-music) Official site: https://whisperweb.site/ Last verified by a human: 2026-10-14 Reviewed by a human at AI Tool Seekers, a hand-reviewed AI tools directory. Every listed tool is verified by a person before publishing. Scores are editorial opinions from a fixed rubric (capability, ease of use, value, delivery on promise) and cannot be bought; there is no pay-to-rank. Directory: https://aitoolseekers.com This review: https://aitoolseekers.com/tools/whisperweb Methodology: https://aitoolseekers.com/how-verification-works ## Verdict WhisperWeb is the right pick for anyone who needs free, private transcription and cannot afford to send audio to a cloud server. The hard ceiling of 200 MB and 20 minutes per file, plus the fact that real-phone transcription is explicitly unverified by the developer, makes it a poor fit for heavy or mobile workflows. If you need more than occasional short-file transcription, Tremigos handles longer files and adds translation and dubbing in a single workflow. Best for: Privacy-conscious users needing free, offline-capable transcription of short files Not for: Anyone transcribing files longer than 20 minutes or working on a mobile device ## Overview WhisperWeb (whisperweb.site) is a browser-based wrapper around OpenAI's open-source Whisper speech recognition model. The entire transcription pipeline runs on your device: no audio is sent to a server, no account is created, and no payment is required. You drop a file, the model runs locally, and you export TXT, SRT, or VTT. That is the whole product. The tool supports a range of audio and video formats including MP3, WAV, M4A, MP4, WebM, MOV, OGG, OPUS, AAC, and FLAC, up to 200 MB or 20 minutes per file. Three Whisper model sizes are available: tiny, base (the default), and small. Larger models produce better accuracy but need more device memory, and the developer explicitly recommends testing with a short clip before committing a long file. The first use downloads the model to the browser cache; subsequent sessions can reuse it without a fresh download. What separates whisperweb.site from the growing crowd of Whisper-based tools is its subtitle workflow. Beyond plain transcription, it generates timed SRT and VTT subtitle files, includes a subtitle format converter (SRT to VTT and back), and strips subtitles to plain text. That makes it genuinely useful for video editors and YouTube creators who need caption files, not just a raw transcript. The developer is candid about limits: full transcription has been tested on desktop Chrome on macOS, and a responsive layout does not mean mobile transcription has been verified. GPU acceleration is used where available, with a CPU fallback. Larger files take longer, and the tool warns users to start with a short, clear recording before tackling anything demanding. ## Who should use it WhisperWeb fits individuals who occasionally need to transcribe interviews, lecture recordings, podcast clips, or short video files and cannot or will not send that audio to a cloud service. A journalist transcribing a sensitive source recording, a student turning a 15-minute lecture clip into notes, or a video editor who needs a quick SRT file without paying for a subscription will all find it does exactly what they need at zero cost. Anyone with files longer than 20 minutes, a heavy weekly transcription workload, or a need for speaker diarization should look elsewhere. Tremigos, which is already in this directory, handles longer files, adds translation and dubbing across 60-plus languages, and is built for teams rather than solo one-off jobs. For users who want cloud-backed transcription with no file-size anxiety and do not mind a paid plan, Tremigos is the more capable option. ## FAQ Q: Is WhisperWeb really free, or is there a paid tier I'm missing? A: WhisperWeb at whisperweb.site is entirely free with no paid tier. The only constraint is the 200 MB and 20-minute file limit per transcription. No credit card or account is required. Q: Does WhisperWeb upload my audio to a server? A: No. The Whisper model runs inside your browser using WebAssembly and local compute. Your audio file is never sent to a remote server. The only network activity is the one-time model download on first use. Q: What file formats does WhisperWeb accept? A: It accepts MP3, WAV, M4A, MP4, WebM, MOV, OGG, OPUS, AAC, and FLAC files, as well as live microphone recordings. Files must be under 200 MB and 20 minutes. Q: Can I use WhisperWeb on my phone? A: The developer explicitly states that a responsive layout does not mean mobile transcription has been verified. Full transcription has been tested on desktop Chrome on macOS. Mobile users should test with a short clip before relying on it. Q: Which Whisper model size should I choose? A: Base is the default and works for most files. Tiny downloads faster and uses less memory. Small offers better accuracy but needs more RAM. The developer recommends testing a short recording before running a longer file on any model. Q: Can WhisperWeb generate subtitle files for YouTube? A: Yes. It exports SRT and VTT files directly, and the site includes a guide specifically on adding generated subtitles to YouTube videos. It also converts between SRT and VTT formats if you need to switch.