mnc lab

mnc lab · for developers building in healthcare

Frontier AI lab for world models

mEVA is our first model, for our first vertical: healthcare. It is a clinical reasoning model of our own design, in the H‑JEPA family of world models, trained end to end on continuous, outcome-linked patient data. It has multimodal capabilities relevant to healthcare, and plugs into your product through one OpenAI-compatible API.

ReMedQA accuracy
89.56%
cheaper than GPT-6 Astra per output token
5×
token context
1M
HIPAA-compliant cloud
BAA

No training on your data by default · deploys to any major cloud, in-country where local law requires

benchmark

More accurate than GPT-6 Astra, at a fifth of the price

ReMedQA · closed-form accuracy higher is better

mEVA
89.56%
Claude Opus 5.5
89.38%
GPT-6 Astra
87.39%
Best open-weights
85.58%
GPT-5.6 Sol
73.81%

Price per 1M output tokens lower is better

mEVA
$10
Claude Opus 5.5
$20
GPT-5.6 Sol
$20
GPT-6 Astra
$50

5× cheaper than GPT-6 Astra, at $2 in / $10 out per 1M tokens

Benchmark results validated by Harvard Medical School.

ReMedQA rewrites medical exam questions, including the USMLE (the US physicians’ bar exam). Run the benchmark yourself on platform.moonco.one.

where mEVA excels

Real-world care needs a 360° patient view

mEVA analyzes images, video, documents and sensor data in one model, and returns output a care team can act on.

Wounds
& photos

  • Tissue classification
  • Wound area measurement
  • Progression analysis over time

Motion
& video

  • Range of motion
  • Stroke facial asymmetry
  • Mobility assessment

Medical
documents

  • OCR of paper charts
  • Extraction from lab photos
  • Structuring clinical notes

Sensor
data

  • Continuous vital signs
  • Smart wearable telemetry
  • Biometric data streams
clinical areas
Wound careOrthopedicsStroke & neuroDocs & chartsRemote monitoring
use cases

One API for every healthcare product

Across clinical healthcare and consumer wellness: what you can build in each segment, with no medical model to train.

Clinical healthcare

Care delivery & providers

Triage, remote monitoring and follow-up between visits.

  • Pre-visit intake and follow-up triage
  • Deterioration flags for the care team
  • Escalation to a clinician, with the reason attached
get started

Live in three steps, with no model training on your side

  1. Get a key

    Request an invitation to platform.moonco.one, then top up from $10.

  2. Change three lines

    Point the OpenAI SDK at api.moonco.one/v1: the key, the base URL and the model name.

  3. Scale with usage

    Pay $2 / $10 per 1M tokens as you grow, with no minimum.

Even the best healthcare AI, ours included, is not yet capable of replacing clinical judgment. The model is built to be checked by a clinician, not to stand in for one.