Record and summarize meetings.
No bot in the call.
Audicap lives on your computer, not in your meeting: capture microphone and system audio, transcribe locally on your device, export ready-to-use minutes. You listen, Audicap takes notes.
You want to listen, not take notes. Audicap remembers the conversation, so you stay present.
Everything that puts your audio to work.
Bot-free recording, local processing, open exports. You decide when cloud quality joins in.
Bot-free recording on your desktop
Microphone and system audio, without a meeting bot joining the call. Double-click the transcript to jump to that spot in the audio. Full-text search finds any statement in your archive.
Processed locally
Recording, transcription, speaker detection and AI summaries: all local, or via cloud. With Ollama & co. your data stays on your device, GDPR-compliant.
Trained for German, fluent in 25 languages
Your own AI prompts
Minutes, action items or a follow-up email: you decide what the AI makes of the conversation. Unlimited summaries per meeting, with a custom prompt whenever you need one.
Open exports, no lock-in
Fits right into Obsidian, Notion, Docs or your CRM.
Knows when your meeting starts
Audicap detects Zoom, Teams, Meet & co. and offers to record: one click before things kick off. Recording starts only when you say so, never silently.
Knows your calendar
Audicap shows upcoming events on the home screen and reminds you when your meeting starts. Never miss a meeting again. Participants are already set up as speakers.
The app in detail.
The follow-up is the AI's job.
Audicap creates a complete transcript of your conversation. Your prompts turn it into exactly the document you need. The busywork after the meeting is gone.
Detail instead of bullet points
Complete, well-structured notes. Nothing important gets lost.
Your prompts instead of fixed templates
Say in your own words what you want: which points, which structure, which tone. The AI sticks to it.
Zero manual work afterwards
Minutes, action items, follow-up email: defined once per meeting type, applied to any recording.
Summarize the conversation as client minutes in a formal tone. Highlight decisions, budget & timeline, objections and next steps. Flag open points for the proposal.
- Decision: Launch moved to the end of Q3.
- Budget & timeline: Delivery in Q3, assets due week 28.
- Objection: Is the onboarding buffer enough?
- Next step: Send the proposal with the open points.
For the last few percent of accuracy.
Three of the world's leading engines, fully optional. Strong at speaker diarization, billed through your own API key with no Audicap markup.
Universal-3-Pro: 2.3% word error rate on German in the Open ASR Leaderboard. An independent German benchmark found the most consistent speaker attribution in its test.
Scribe v2: 2.3% word error rate on German, ranking second on the multilingual Open ASR Leaderboard, well ahead of Whisper. Speaker separation for up to 32 voices.
Nova-3 transcribes with under 300 ms latency and returns even long recordings in seconds, with a dedicated German model.
Claude, GPT, Gemini and OpenRouter. Or GDPR-compliant with Ollama and any OpenAI-compatible endpoint, from Amazon Bedrock to your own server. No silent backend switching: whether processing runs locally or in the cloud stays visible and under your control.
Try it free. Your meetings belong to you.
Early Access
- Local transcription without limits: Parakeet and Whisper run on your device.
- Prompt library included: your own AI templates for minutes, action items or follow-up emails.
- Meeting detection & calendar: Audicap offers to record when your meeting starts.
- Cloud optional: through your own API key, pay only for usage.
macOS 12+ · Windows 10/11 · one license, two devices
Frequently asked questions.
Am I allowed to record meetings with Audicap?
That depends on context and local law. Audicap is built for transparent use: recordings never start automatically, and a reminder (on by default) asks before every start whether participants have consented.
Does my data stay local?
Local-first: transcription runs entirely on your device, and with Ollama so do AI summaries. Cloud backends are optional and only used if you deliberately configure them with an API key. There is no silent backend switching.
Can I use Audicap in a GDPR-compliant way?
Yes. In local mode, recordings and transcripts never leave your computer, no cloud required. For AI summaries you can use Ollama locally or any OpenAI-compatible endpoint you choose, such as Amazon Bedrock in an EU region or your own server. Data processing stays fully under your control.
How well does Audicap handle German meetings?
Audicap uses models specifically trained and optimized for German, which helps with terminology, names and punctuation. Parakeet V3 Multilingual covers 25 European languages.
Can I adapt summaries to my use case?
Yes. With your own AI prompts you define per meeting type what comes out: client call, interview, 1:1, sales or standup. You decide which points the AI extracts, in which structure and tone. Once defined, a prompt can be applied to any new recording.
Can I import existing recordings?
Yes, via drag-and-drop or file dialog. Audicap transcribes M4A, MP3, WAV, FLAC and more, just like its own recordings. Great for interviews, dictations and old recordings.
Which meeting tools does Audicap work with?
All of them, as long as the audio runs through your computer: Audicap captures system audio and microphone directly on the device, no integration needed. Meeting detection recognizes Zoom, Teams, Google Meet, Webex, FaceTime and Slack.
What does Audicap cost?
Audicap is free during the beta. Just download and try it. It will become a paid product later; beta users will be notified in advance. Local transcription is included without limits, cloud backends run on your own API keys.