What it needs, and what we will not promise yet.
The numbers below are the ones we are confident in today. Processing times are published with the release, measured on named hardware, not before — we do not quote speed from a development machine that is faster or slower than yours.
It runs on the laptop you already have.
Within reason, and we would rather tell you the limits now than refund you later.
- Runs
- Clinical notes
- Operating system
- Windows 10 or 11, 64-bit
- Windows 10 or 11, 64-bit
- Processor
- 4 cores with AVX2 (roughly 2013 onward)
- 8 cores recommended
- Memory
- 8 GB
- 16 GB
- Disk
- 5 GB
- 8 GB
- Graphics card
- Not required
- Not required; a Vulkan GPU makes notes faster and enables non-English transcription
On 8 GB we have to run a smaller language model, and in our testing that model attributed the clinician's own words to the client — in quotation marks. The larger model that needs 16 GB got every attribution right on the same transcript. Putting the clinician's words in the client's mouth is a factual error in a medical record, not a wording preference, so we do not ship clinical templates on the 8 GB tier at all. An 8 GB machine still records, transcribes, separates speakers and produces plain summaries. If your laptop has 8 GB and you need SOAP notes, this is not the right purchase yet, and we would rather lose the sale than take it.
We develop on a deliberately slow machine — an eight-year-old CPU-only workstation — because that is closer to a working professional's office laptop than a developer's desktop. Every design decision in this product came out of measuring it there. Transcribing during the session rather than after it, for instance, is not a feature we thought of first; it is what the benchmark on that machine forced.
Processing times and the raw benchmark data are published with the release, measured on that machine and on a current laptop, so you can find hardware like yours on the list rather than trust an average. Spokenroom tells you which performance tier your computer is on during setup instead of hiding it.
A macOS version is planned. Join the waitlist and you will hear about it first.
Three tiers of machine, each answering only its own question.
- Minimum tier — an 8 to 16 GB laptop with an AVX2 CPU and no GPU. This is the only machine whose numbers go into the requirements table above.
- Recommended tier — a current desktop CPU with no GPU acceleration enabled. Tells us what a modern office machine does.
- GPU tier — used for deciding which optional accelerated packages are worth shipping. Its numbers never appear in the requirements table.
Memory figures do not depend on the machine and are already fixed: the smaller note model peaks at about 2.7 GB and the clinical model at about 4.8 GB, on top of roughly 1.7 GB for transcription. That is why 8 GB runs transcription comfortably and 16 GB is required for clinical templates.