Local meeting transcription that never leaves your computer

Daisy does local meeting transcription: the audio is captured on your device, turned into a transcript on your device, and saved as files you own. The recording never has to leave the device it was made on.

That covers the whole core loop: recording, the transcript, speaker labels, and full-text search. The cloud enters the picture only if you choose a cloud AI provider for summaries, spelled out below.

What "local" actually means here

Local means the work happens on your own hardware and the output lands on your own storage. Daisy has no account to create. One license covers three devices; it activates online and then works offline between occasional license checks.

The files are open and yours: the audio, and the transcript as plain markdown. You can search them, back them up, feed them to whatever tool you like, or delete them.

How Daisy transcribes on-device

Recording, transcription, speaker identification (diarization), and search all run locally.

Speaker labels are always produced on-device, in every configuration, with no cloud key involved.

When you stop a meeting, Daisy finishes the transcript locally on ordinary hardware. That transcript is the canonical record; everything else is layered on top of it.

Live captions and AI

Live captions show words on screen as people talk.

Which speech model does the work depends on your platform, and all of them run locally:

• Windows and Linux — an English model bundled into the app, with nothing to download.
• Recent macOS, iPhone, and iPad — Apple's own on-device speech engine.
• Whisper — downloadable models in several sizes on desktop, for more languages and for older Macs.

Live captions and the finished transcript are both produced locally, on every device.

Summaries, Q&A, and analysis need a language model. You have four options:

• On-device AI — Daisy's built-in model writes summaries and answers questions on the device, with no key. It needs recent hardware: iPhone 15 Pro or later, or a computer with a graphics card.
• A local LLM via Ollama or LM Studio — fully private, nothing leaves the machine.
• A cloud key you bring (OpenAI, Anthropic, Groq) if you'd rather use a hosted model.
• None at all, on desktop — Daisy hands you the transcript plus a ready-made copy-paste prompt, so you can run it wherever you like.

On a computer, the desktop app uses its built-in model, a local AI server, or your own key.

Private by default, your choice to extend.

The trade-offs

• Keeping up with live speech takes real compute. Most machines will support real-time transcription. Some slower machines won't show the real-time transcriptions and will only do the transcription once the session is completed.
• Good AI requires good resources. The on-device AI's summaries and Q&A will be good enough for most people. But we give you options, and if you want to run something big we give built-in integrations to handle most requirements.
• You are responsible for backups. Your profile is stored where you choose. If it's not backed up, that's on you. Recordings sync from your phone to your computer. Think "Shroud". IYKYK.
• Daisy is an app you run, and no bot joins your calls. You record the audio in the room or on your device.

If those trades read as features rather than costs, you're the audience.

Who this is for

Daisy is for privacy and local-first folks who'd rather run a local LLM than hand a transcript to a SaaS. This includes anyone under a policy — legal, clinical, regulated, or just personal — that says recordings don't go to someone else's cloud.

On price: free tools tend to pay for themselves with your data. Daisy charges a modest fee instead, so your meetings stay yours. Our apps are also signed by Apple and Microsoft and license fees cover such costs.

Daisy runs on macOS, Windows, and Linux, with Daisy for iPhone and iPad. The 30-day trial lets you confirm the local loop on your own hardware before deciding anything.