In development for macOS
The local-first qualitative analysis workspace for Mac.
Transcribe, code, and analyze your research interviews — entirely on your Mac. No cloud upload, no per-seat subscription, no wondering where your participants' recordings ended up.
What it looks like
A workspace built around the transcript, not the software.
Sources, transcript, and codebook live in one window — no exporting to spreadsheets to see what you have.
Three commitments
Built for research, not repurposed for it.
Truly local
Transcription, speaker separation, and AI analysis run on your Mac — not in a data center. Sondera works with no network connection at all, and shows a live indicator whenever it does talk to the network, so "local" is something you can watch, not a line in a privacy policy. Built for the realities of IRB review and participant confidentiality.
AI suggests, you decide
Sondera proposes codes, clusters emerging themes, and surfaces related quotes by meaning, not just keyword — every suggestion sits in a pending state until you accept it. Every accept, reject, and edit is written to an audit trail you can export, so when a reviewer asks what the AI actually did, you have an answer.
One-time purchase
$179, once. Early access $129, students $79. No annual renewal — and this is a promise, not a hedge: no future version of Sondera will lock you out of a project you built in an older one.
How it works
Four steps, one file, zero uploads.
Import
Bring in audio or video interviews directly. Nothing is queued to a server to get started.
Transcribe
A local, Whisper-class model transcribes and separates speakers automatically, on-device.
Code
Apply codes with a keyboard-driven flow, build a code tree, and pull every quote for a code in one view.
Analyze
AI-assisted theme clustering, with matrices and quote tables that export to Word and Excel.
Why not NVivo?
The honest comparison.
No mainstream QDA tool runs its AI features on your machine — yours goes to a vendor's cloud the moment you turn AI on. That's the gap Sondera is built to close.
Scroll to compare →
| Sondera | NVivo | MAXQDA | ATLAS.ti | |
|---|---|---|---|---|
| License | $179 once early $129 · student $79 |
Academic subscription $295–595 / yr |
Subscription ~$253 / yr |
Subscription |
| AI features | Included | +$250 / yr add-on | Extra cost | — |
| Where AI runs | On your Mac, always | Cloud (OpenAI) | Cloud | Cloud |
| Heard in the field | —, not yet launched | "Upgraded and my coding was gone" | — | "10–15 seconds per code click" |
Questions
Frequently asked
Does my data ever leave my Mac?
No. Transcription, speaker separation, and AI analysis all run locally, on-device. Sondera shows a live indicator whenever it makes a network request, so "local" isn't something you have to take our word for — it's something you can watch. The app works fully with no network connection at all.
What about IRB compliance?
Because recordings and transcripts never leave your Mac by default, Sondera fits cleanly into protocols that require participant data to stay on a single, non-networked device. It doesn't replace your IRB's judgment — but it removes "did this vendor's cloud see our data" from the list of questions you have to answer.
Which Macs are supported?
Apple Silicon (M1 or later), macOS 14 or later, at launch. We're asking about Intel Macs, Windows, and Linux in the waitlist form below to see how much demand there is — nothing is ruled out, but the local AI models we rely on need Apple Silicon to run well today.
When will it launch?
Beta this fall (fall 2026). Waitlist members get priority access and first word on pricing.
Do you offer student pricing?
Yes — $79 with a valid .edu (or equivalent institutional) email address, honored at launch.
Can I import from NVivo?
REFI-QDA (QDPX) import is on the roadmap. It's the closest thing our field has to a portable project format, and we intend to support it.
Get early access
Be first to know when Sondera is ready.
Beta opens fall 2026. Waitlist members get priority access and lock in early-bird pricing before the public launch.