Skip to main content
Mother Nature AI
Developer Platform

Build with
health intelligence

REST API access to AskMN v3, Sylvia 2.1, NutriGen 1.0, and Genlyy 1.3. Purpose-built health AI models with structured outputs, evidence citations, and HIPAA-conscious infrastructure.

SFT + RLHF · RAG over 50M+ papers · Structured JSON I/O · HIPAA BAA available

Request API Access View Model Specs
REST + JSON
HIPAA BAA available
P50 < 1.2s
API key auth

Available Models

Four specialized health AI models

Each model is domain-specific. Request access for the models relevant to your use case.

Generally Available

AskMN v3

Herbal medicine · Toxicology · General health

The primary health and herbal intelligence model. Reasons across 3,000+ medicinal plants, drug-herb interactions, toxicology risk, and integrative health protocols with evidence citations.

Request access
Generally Available

Sylvia 2.1

Mental health · Crisis detection · CBT/DBT

Specialized mental health model fine-tuned on CBT, DBT, trauma-informed care, and psychiatric literature. Includes 98.2%-sensitivity crisis detection and safe messaging protocol activation.

Request access
Beta

NutriGen 1.0

Nutrigenomics · Dietary optimization

Multi-modal nutrigenomics engine. Fuses SNP data, dietary logs, and micronutrient panels to generate genotype-specific nutrition protocols calibrated to 200+ genetic loci.

Request access
Invite Only

Genlyy 1.3

Genomic variant interpretation · PGx

Clinical-grade genomic variant interpretation. Processes VCF/SNP arrays and returns ACMG/AMP-classified variant reports across 4,800+ loci with FDA pharmacogenomics cross-reference.

Request access

API access is granted on a per-project basis after a brief review of your use case, technical setup, and data handling practices. Access decisions typically take 1–3 business days. Genlyy 1.3 requires additional partner vetting.

Authentication

API key authentication

The Mother Nature AI API uses bearer token authentication. Include your API key in the Authorization header. Keys are project-scoped with configurable model-level permissions and IP allowlists.

  • Project-scoped keys — separate keys per environment (dev / staging / prod)
  • Keys can be restricted to specific model families and IP allowlists
  • Full audit logs of all API key activity in the developer dashboard
  • Zero-downtime key rotation with dual-key overlap support
  • OAuth 2.0 / OIDC enterprise SSO integration available
Authentication
# Set your API key as an environment variable
export ASKMN_API_KEY="mnai_sk_live_xxxxxxxxxxxx"

# Include in every request header
Authorization: Bearer $ASKMN_API_KEY

# Verify your key
curl https://api.askmn.ai/v1/models \
  -H "Authorization: Bearer $ASKMN_API_KEY"

# Response
{
  "models": [
    { "id": "askmn-v3", "status": "available" },
    { "id": "sylvia-2.1", "status": "available" },
    { "id": "nutrigen-1.0", "status": "beta" },
    { "id": "genlyy-1.3", "status": "invite_only" }
  ]
}
200 OK — Response body
{
  "id": "chat_01JK8X2F4M9NV3P",
  "model": "askmn-v3",
  "content": "Ashwagandha (Withania somnifera) contains withanolide glycosides that may modestly inhibit CYP2C9 — the primary enzyme responsible for warfarin (S-warfarin) metabolism. Co-administration carries a theoretical risk of elevated warfarin plasma levels and increased bleeding risk [PMID: 28601062, 31394563].",
  "structured_findings": {
    "interaction_severity": "moderate",
    "mechanism": "CYP2C9 inhibition (theoretical, in vitro evidence)",
    "clinical_recommendation": "If patient proceeds, monitor INR within 1–2 weeks of initiating ashwagandha. Dose reduction of warfarin may be required.",
    "contraindication_flags": [
      {
        "type": "drug_herb_interaction",
        "compound_a": "warfarin",
        "compound_b": "ashwagandha",
        "severity": "moderate",
        "evidence_level": "in_vitro_theoretical"
      }
    ]
  },
  "citations": [
    {
      "pmid": "28601062",
      "title": "Withania somnifera CYP inhibition in human liver microsomes",
      "journal": "Phytomedicine"
    },
    {
      "pmid": "31394563",
      "title": "Herb-drug interactions in anticoagulant therapy",
      "journal": "British Journal of Clinical Pharmacology"
    }
  ],
  "confidence_level": "moderate",
  "escalation_flag": false,
  "usage": { "input_tokens": 94, "output_tokens": 284 }
}

Structured Outputs

Machine-readable health intelligence

Every model returns structured JSON alongside conversational text — making it straightforward to build clinical features without parsing free-form responses.

Interaction severity classification

Structured compound-pair objects with severity level, mechanism of action, and evidence grade.

Evidence citations

Every clinical assertion includes PMID references and journal attribution.

Contraindication flags

Machine-readable flag objects for drug-herb interactions, toxicology risks, and crisis signals.

Confidence level

Model-assessed confidence (high / moderate / low) based on evidence density and case complexity.

Escalation flag

Boolean signal for red-flag situations requiring immediate clinical or crisis intervention.

Use Cases

What builders are creating

Integrative Health Apps

Embed AskMN v3 as an herbal and integrative health reasoning layer. Users ask about supplements, herbal combinations, and plant-drug interactions and receive evidence-cited, toxicology-aware guidance.

Mental Wellness Platforms

Power CBT and mindfulness apps with Sylvia 2.1. Crisis detection, psychoeducation delivery, and PHQ-9/GAD-7 severity classification built into every interaction.

Nutrition & Wellness SaaS

Layer NutriGen 1.0 on food logging or supplement platforms to deliver genotype-specific nutrition protocols. Connects seamlessly with Genlyy for full genomic context.

Genetic Testing Labs

Use Genlyy 1.3 as an AI interpretation layer for raw DNA test outputs. Turn VCF files into clinical-grade patient reports with medication metabolism profiles.

EHR & Clinical Systems

Integrate AskMN v3 summarization and analysis endpoints to generate AI-assisted clinical notes, surface herb-drug interaction flags, and support integrative medicine workflows.

Research Platforms

Access model APIs for population-level health research, genomic cohort analysis, and clinical decision support development under IRB-governed data use agreements.

SDKs

Official client libraries

Python

Coming Soon

In Development
pip install askmn

Full-featured Python SDK with type hints, async support, streaming, and automatic retry. Python 3.9+ compatible.

JavaScript / TypeScript

Coming Soon

In Development
npm install @askmn/sdk

Isomorphic SDK for Node.js and browser. Full TypeScript types, ESM + CJS, streaming via async iterators.

REST (Direct)

API v1

Available
No install required

Direct REST API access via HTTP. Contact us for API key and documentation.

Compliance & Security

Infrastructure you can build on

HIPAA BAA

Business Associate Agreements available on approved plans. PHI processed in HIPAA-eligible isolated compute environments.

SOC 2 Type II & HIPAA

SOC 2 Type II certified and HIPAA compliant. Security controls documentation available to partners under NDA.

Data Residency

US inference by default. EU data residency available for GDPR-governed deployments on enterprise agreements.

Audit Logging

Full audit trail of all API calls and model outputs. Exportable to your SIEM. Configurable retention policies.

Ready to build?

Submit a request for API access. After a brief review of your use case and data practices, we'll send your credentials within 1–3 business days.

Request API Access View Model Specs