Terms
Use prelaunch access with your own local audio policy.
Plainsong launch materials describe a local-first desktop workflow for dictation, meeting capture, transcription, review, and export. Public release details are still being finalized.
You are responsible for getting consent before recording conversations and for following workplace, platform, and regional recording rules that apply to your use.
Plainsong is not currently offered as a checkout product through this site. For terms questions, use the Plainsong contact page.
Treat prelaunch access as evaluation
Prelaunch access should be treated as product evaluation, not a guarantee that every planned workflow is available. Test with audio you are allowed to process, keep your own backup of important records, and avoid relying on draft behavior for legal, medical, HR, or customer commitments without independent review.
The safest way to evaluate the product is to use low-risk sample audio first, then move toward real workflows only after you understand the capture, review, and export limits. Plainsong is meant to help turn speech into usable writing, but the person recording still owns consent, retention, review, and final use of the transcript.
Output is a draft until a person reviews it
If a workflow has higher stakes, treat Plainsong output as a draft record until a person reviews it. A transcript can speed up follow-up, but it should not replace the judgment needed for commitments, approvals, private records, or any situation where an inaccurate note could cause harm.
Launch materials may change before release
Product notes, screenshots, release language, and feature examples should be read as launch materials until public release details are final. If the product changes before release, the current site should be updated so the terms, privacy notes, and subprocessor disclosure describe the same workflow.
These terms are intentionally written for a prelaunch product. They should stay plain until public access, support obligations, update cadence, and any hosted features are final. Until then, assume the safest path is careful evaluation, local backups, and human review before relying on transcript output.
Last updated