GhostScribe transcribes audio and video with OpenAI’s Whisper model running in your browser — to plain text, SRT, or VTT — so the recording never leaves your device. Free, no account, no minutes quota.
Otter.ai is a meeting assistant: it joins your calls, transcribes them live, labels who said what, and turns the conversation into summaries your team can search and share. To do that, your audio is processed in Otter’s cloud and the transcripts live in your Otter account.
GhostScribe does one narrower job, differently. You drop a recording you already have — an interview, a voice memo, a lecture, a video — and Whisper, OpenAI’s open speech-recognition model, transcribes it on your own device via WebGPU or WebAssembly. The audio is never uploaded, there is no account for it to live in, and the result downloads as plain text or as timestamped SRT or VTT subtitles.
Otter is a genuinely capable product for meetings. It can join Zoom, Google Meet, and Microsoft Teams calls on its own, transcribe in real time, identify speakers, and produce AI summaries and an AI Chat over your meetings.
It is also collaborative: transcripts and summaries live in a shared workspace you can search and share with the people you choose. Because the heavy lifting happens in Otter’s cloud, none of that depends on how powerful your own laptop is. If your work runs on meetings, that combination is hard to match.
The recording never leaves your device
The audio is decoded and transcribed inside your browser tab. The only downloads are the speech model and the audio decoder — self-hosted by GhostX and cached after first use — never your file going the other way.
Free, with no minutes quota
There’s no monthly minute allowance and no limit on how many files you import. Your own device does the work, so there is nothing to meter — the only ceiling is file size (500 MB on desktop, 150 MB on phones), because the browser holds the file in memory.
No account, no transcript archive
Nothing to sign up for, and no copy of your transcript stored anywhere but the file you download. For privileged, medical, HR, or source interviews, that absence is the point.
Subtitles ready to use
Export plain text, SubRip (.srt), or WebVTT (.vtt) with per-segment timestamps — ready for a video editor or player. Works on video files too: GhostVideo extracts the audio track locally first.
Mute what shouldn’t be heard
Because it all runs locally, the same engine powers GhostAudio’s speech redaction, which finds spoken phone numbers, emails, card numbers, and names and mutes those segments — also without uploading.
| Feature | GhostScribe | Otter.ai |
|---|---|---|
| Where audio is processed | On your device (Whisper in the browser) | In Otter’s cloud |
| Transcription languages | Auto-detect + 13 common languages in the menu | English, Spanish, French, German, Japanese, Chinese |
| Account required | No | Yes |
| Price | Free | Free Basic; Pro from $8.33/user/mo billed annually |
| Usage limits | No minute quota; files up to 500 MB (150 MB on phones) | Basic: 300 min/month, 30 min per conversation, 3 lifetime file imports |
| Live meeting capture | No — recordings you already have | Yes — joins Zoom, Meet, and Teams |
| Speaker identification | No | Yes |
| AI summaries and chat | No | Yes |
| Team sharing and comments | No — you get a file | Yes |
| Subtitle export (SRT / VTT) | Yes, free | SRT on paid plans; Basic exports TXT and MP3 |
Keep Otter if you need a live meeting assistant: something that joins the call for you, transcribes as people talk, tells you who said what, and hands the team a summary afterward. GhostScribe does none of that — it transcribes a file you already have, with no speaker labels, no summaries, and no shared workspace.
Otter’s cloud will also often be faster and more accurate on long, crosstalk-heavy English meetings. GhostScribe runs Whisper’s base model on your hardware, which is good on clear speech but slower on older devices and less accurate than large cloud models on noisy audio. Choose GhostScribe when the recording is sensitive enough that uploading it is the problem.
Is GhostScribe really free?
Yes. There’s no account, no plan, and no minute allowance. The transcription runs on your own device, so there’s no per-minute server cost to pass on. The practical limits are your hardware’s speed and a file-size ceiling of 500 MB on desktop (150 MB on phones).
Is my audio uploaded anywhere?
No. The Whisper model downloads to your browser once (about 140 MB for the multilingual model, about 95 MB for the fast English-only one) and is cached. Your recording is decoded and transcribed locally and never sent to a server.
Can it transcribe a live Zoom or Teams meeting?
No. GhostScribe transcribes recordings you already have. If your meeting tool lets you save a recording, drop that file in afterward.
Does it label speakers?
No — the output is a single transcript with timestamps, not a speaker-by-speaker script. Speaker identification is one of the things Otter does that GhostScribe doesn’t.
What languages does it support?
The default multilingual model can auto-detect the spoken language, or you can pick from a list of common languages including English, Spanish, French, German, Portuguese, Chinese, Japanese, Arabic, and Hindi. A faster English-only model is also available.
Can I get subtitles for a video?
Yes. Drop the video into GhostVideo’s transcribe tool and choose SRT or VTT. The audio track is extracted and transcribed on your device, and you get a timestamped subtitle file.
Whisper on your own device — text, SRT, or VTT. No upload, no account, no minutes quota.
Transcribe a recording free