Platform API · public beta
REST and streaming API for Lor-1. TypeScript and Python SDKs. Usage-based pricing, no commercial call required. Here is how to integrate Lor-1 into your clinical product in an afternoon.
Lorraine Platform is in public beta today. It is a REST and streaming API for Lor-1 — the same clinical foundation model that powers Chat and Learn, exposed with the familiar OpenAI-shape completions interface and first-class citation support.
Who it is for
Healthtech builders who need clinical reasoning and SA guideline coverage in their own product: EHR add-ons, triage assistants, documentation tools, clinical search. If you are building software that a clinician uses to make decisions, and you need grounded output rather than generic chat, Platform is for you.
Pricing and limits
Free tier: 10,000 tokens a month, 60 RPM, no credit card. Developer tier: R0.08 per 1,000 output tokens, 10,000 RPM baseline, email support, signed DPA on request. Enterprise: dedicated capacity, 99.9% uptime SLA, contractual no-training clause, SSO, audit logs. Metered, not seat-based. Scale from zero to production without a commercial call.
SDKs
TypeScript and Python SDKs are on npm and PyPI today (@lorraine/sdk and lorraine). Both mirror the OpenAI client shape, with two Lorraine-specific extras: cite_sources returns inline citations with full metadata, and structured_output supports typed JSON extraction for EHR-grade ingestion.
Docs live at docs.uselorraine.co.za. If you build something interesting on the beta, we want to hear about it — platform@uselorraine.co.za.
/written-by
Lorraine Team
Engineering
/keep-reading
All articles- Announcements
Announcing Lorraine Chat
Clinical AI for practising South African clinicians, generally available today. Evidence-grounded answers, SA guideline coverage, and HPCSA CPD tracking — built from a year of clinician interviews.
- Announcements
Introducing Lorraine: A New Chapter in South African Medical Education
Lorraine launches as the CMSA-focused learning platform built for South African doctors, with local-first content and adaptive learning.