Lor-1 APIv1.0 · access by form

The Lor-1 API/clinical-grade AI.

A REST + streaming API for the South African clinical foundation model. Differentials, guideline retrieval, clinical summarisation, and structured extraction — with the safety scaffolding clinical software needs.

View pricing

/request

curl https://api.uselorraine.co.za/v1/chat/completions \
  -H "Authorization: Bearer $LORRAINE_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{
    "model": "lor-1",
    "messages": [
      {"role": "system", "content": "You are a clinical reasoning assistant."},
      {"role": "user", "content": "Differential for acute chest pain in a 45yo smoker?"}
    ],
    "cite_sources": true
  }'

Surface: REST + streaming
Auth: Bearer API key
Compat: OpenAI-shaped
Latency: INT4 + speculative

Capabilities

What Lor-1 does well.

Four primitives, all benchmark-evaluated on SA clinical questions before they ship. Compose them into the workflow your users already know.

Primary

Differential generation

Structured, ranked differentials with pre-test probability reasoning and red-flag escalation.

Guideline retrieval

Hybrid RAG over SA clinical guidelines + international evidence bases.

Clinical summarisation

Turn a long handover note into a structured clerking summary.

Structured extraction

Extract medications, labs, diagnoses as typed JSON — EHR-ready.

Model cardProduction

Lor-1/clinical reasoning.

A clinical foundation model calibrated to reason inside South African clinical conventions — drug names, formulary, guidelines, and referral pathways.

Open the model card