Why not local Whisper — and why we went local anyway

When Lven launched, we deliberately chose server-side transcription. The Whisper large-v3 model was 1.5 GB, slow on every consumer device we benchmarked, and drained battery hard on phones. Sending audio to dedicated hardware behind a TLS connection produced text in under a second — fast enough that nobody noticed the round trip.

What changed: a compact speech model runs fully on-device inside a ~200 MB install footprint and matches the accuracy band most users actually need. On a five-year-old laptop it transcribes faster than people can finish speaking — fast enough that the on-device build became the desktop default.

Once a credible local model existed, the cloud trade-off became indefensible. Sending audio to a server has two unavoidable costs: privacy exposure (the audio crosses infrastructure you don't control) and network dependence (a slow connection means slow transcription). Local removes both.

Lven Instant ships on Windows, Linux, and Android today with the on-device quantised model, and is rolling out to macOS and iPhone. The server-side product, Lven Cloud, is the right choice for hardware below Instant's CPU floor, for platforms the on-device engine hasn't reached yet, or for anyone who wants the highest-accuracy cloud engine. But everywhere Instant ships, on-device is the default — because it can be.