Common questions

Download the app and sign in with Google, Apple, or email. Cloud sync is enabled automatically — no additional setup or API keys to configure. You're ready to record in under a minute.

Live recording works on any iOS or Android device using the built-in mic. For uploads, Tapelog accepts mp3, m4a, wav, and mp4. The processing pipeline runs server-side so quality and length don't matter — even long recordings work fine.

You can record without an internet connection. The audio is saved locally and will be queued for AI processing once you're back online. Existing processed recordings are accessible offline too.

Transcription supports over 50 languages. The AI analysis (summaries, action items, decisions) currently works best in English, with strong support for Spanish, French, German, Portuguese, and Japanese. Real-time translation lets you view captions in a different language from the one being spoken.

Yes. Audio is stored in a private cloud bucket scoped to your user ID with row-level security — only you can access your recordings. API keys are stored server-side as secrets and never appear in the app or on your device. We don't sell your data and we don't train AI models on your recordings.

Audio is uploaded to your private cloud storage, then processed server-side in a secure environment. Transcription runs against your audio file, the transcript is passed to the AI for analysis, and the results are written back to your account. At no point does audio pass through any third-party beyond the transcription and AI services — both of which process your data under their enterprise privacy terms.

Yes. Individual recordings can be deleted from the app at any time. To delete your entire account and all associated data, contact us at hello@tapelog.ai and we'll permanently remove everything within 30 days.

Laws around recording consent vary by country and state. Many jurisdictions require all-party consent. You are responsible for ensuring you have the legal right to record any conversation. Tapelog is designed for personal note-taking, interviews with willing participants, and meetings where all attendees are aware of the recording.

Live captions use the device's built-in speech recognition (iOS SFSpeechRecognizer / Android SpeechRecognizer) — no extra API call. Text appears in real time as you speak. Multiple speakers are colour-coded using turn detection. Toggle on in Settings → Accessibility. Real-time translation can be enabled alongside captions.

Real-time translation runs on the caption stream. As words are recognised, they're translated into your chosen language and displayed in the captions panel. Currently supports translation into English, Spanish, French, German, Japanese, Portuguese, Italian, and Dutch. More languages coming. Enable it in Settings alongside live captions.

Voice journal mode is a dedicated recording type for personal reflection. When you record in journal mode, the AI doesn't extract action items or decisions. Instead it analyses your entry for mood (scored 1–10), surfaces patterns in your thinking, identifies your core topic, asks follow-up questions to deepen your reflection, and offers an affirmation. Your mood scores are tracked on a timeline so you can see emotional patterns over weeks and months.

The AI automatically detects explicit promises in your recordings — "I'll send that by Friday", "let me get back to you on that." These are extracted with detected due dates, sorted by open/overdue/done, and you can snooze or mark them complete. A Friday review nudge reminds you to clear the week's open commitments.

Type a meeting topic and optionally select attendees from your People profiles. Tapelog searches all your past recordings for relevant context, open actions, and commitments, then writes a brief covering: what's been discussed before, what's still unresolved, key questions to ask, and what to watch out for. Tap "Generate brief" at least a few minutes before you walk in.

The first time you rename a speaker in a recording (e.g. "Speaker 1" → "Sarah"), Tapelog stores a voice profile for that person. In every future recording, Tapelog compares detected speakers against stored profiles and automatically labels anyone it recognises — no manual renaming needed. The system improves with every conversation you have with the same people.

Health scores run entirely on-device with no extra API call. Tapelog analyses 5 dimensions from data already extracted: speaker balance (who dominated), clarity (real decisions made), actionability (action items captured), focus (output relative to meeting duration), and energy (tone from word analysis). You get an A–F grade, specific flags like "One speaker talked 80% of the time", and one improvement tip targeting the weakest dimension.

Pattern memory scans your recording titles, summaries, action items, tags, and transcript excerpts to find topics that keep recurring across conversations. It surfaces recurring blockers (topics appearing 3+ times with words like "delay", "blocked", or "issue"), shows a timeline of when each topic first emerged, and marks topics as resolved once they haven't appeared for 14 days. Runs entirely on-device — no AI call needed.

Tapelog exports to: PDF, Markdown, JSON, HTML, CSV, Notion-optimised Markdown, Obsidian (with YAML frontmatter), and plain text. All 8 formats are available on the Free plan. The export includes the title, summary, action items, decisions, key moments, tags, and full transcript.

The free plan includes up to 5 recordings per month with full AI processing (summaries, action items, decisions, speaker insights, key moments), live captions, voice journal mode, and export to all 8 formats. No credit card required to sign up.

When you start a Pro or Team plan, you get 7 days completely free with no credit card required upfront. At the end of 7 days you can choose to continue (and add a payment method) or let it lapse back to the Free plan — your data stays intact either way.

Pro monthly is $15/month. Pro yearly works out to $12/month ($144 billed annually) — a saving of $36 per year. Both include the same features. You can switch between monthly and yearly at any time from your account settings.

Yes. Cancel from Settings at any time. Cancellation takes effect at the end of your current billing period — you keep Pro access until then. We don't offer partial refunds except where required by law, but if you believe you were charged in error contact us at hello@tapelog.ai within 30 days.

We're actively building it. The bot joins your online meetings automatically, records audio server-side, and runs it through the same AI pipeline — no phone held up, no app kept open. Expected in the Pro plan in 2026. Sign up to get notified when it launches.

Real-time translation is in active development and will be available to Pro subscribers. The captions infrastructure (which translation builds on) is already live. We're currently expanding language support and optimising for latency. Expected rollout in 2026.

The developer API is targeting Q4 2026. It will expose all core features — transcription, AI analysis, speaker data, action items, commitments, health scores, and pattern memory — via a clean REST API. Join the waitlist on the homepage to get early access.