API Documentation

Integrate MolForge into your workflow with our REST API

New · AI Foundry

Foundation models
for drug discovery.

Deploy MoLFormer-XL, Uni-Mol, and ChemCrow + Llama-3 as a managed pipeline. Fine-tune on your proprietary HTS data with full IP ownership. Pharma-tier pricing with VPC isolation and SOC 2 / HIPAA compliance.

Explore AI Foundry Train Your Model

POST /foundry/admet
MoLFormer-XL · 9 endpoints · 82ms

POST /foundry/dock
Uni-Mol · top-5 poses · ΔG + PLIF

POST /foundry/agent
ChemCrow · de-novo design loop

POST /foundry/jobs
Fine-tune or pre-train on your data

Quick Start

Get up and running with the MolForge API in two simple steps.

Get Your API Key

Navigate to your Settings page and generate a new API key. Keys are prefixed with mf_ for easy identification.

Make Your First Request

Use the API key in the X-API-Key header to authenticate your requests.

curl

curl -X POST https://api.molforge.ai/v1/molecules/analyze \
  -H "X-API-Key: mf_your_api_key_here" \
  -H "Content-Type: application/json" \
  -d '{"smiles": "CC(=O)Oc1ccccc1C(=O)O"}'

Response

{
  "molecule_id": "mol_a1b2c3d4e5f6g7h8",
  "smiles": "CC(=O)Oc1ccccc1C(=O)O",
  "name": "Aspirin",
  "properties": {
    "molecular_weight": 180.16,
    "logP": 1.31,
    "hbd": 1,
    "hba": 4,
    "tpsa": 63.60,
    "lipinski_violations": 0
  },
  "admet": {
    "absorption": "High",
    "bbb_permeation": false,
    "cyp_inhibition": false,
    "hepatotoxicity": "Low",
    "ld50": 200
  },
  "credits_used": 1
}

Authentication

All API requests require authentication. Choose the method that fits your workflow.

API Key Header

Pass your API key in the X-API-Key header. Recommended for server-to-server calls.

X-API-Key: mf_your_api_key_here

Bearer Token

Alternatively, use the standard Authorization header with a Bearer prefix.

Authorization: Bearer mf_your_api_key_here

Generate and manage your API keys on the Settings page.

PDF Reports

Generate detailed molecular analysis PDFs from the dashboard. Each report includes 12 sections: executive summary, physicochemical properties, drug-likeness analysis, structural alerts, bioactivity predictions, synthesis feasibility, commercial availability, literature references, and more.

How to use: Analyze any molecule on the Dashboard, then click the PDF Report button.

Limits: Academic — 10/mo | Pro — 100/mo | Team — Unlimited | API — 50 credits/report ($1)

Rate Limits

Tier	Requests / min	Requests / day
Free	—	—
Academic	30	1,000
Pro	60	10,000
Team	120	50,000
Enterprise	Custom	Custom

API Endpoints

Complete reference for all available endpoints. Click any endpoint to see parameters and examples.

Molecules

POST /v1/molecules/analyze — Analyze a molecule (properties, ADMET, bioactivity, synthesis)

Parameter	Type	Required	Description
`smiles`	string	Required	SMILES string of the molecule
`properties`	array	Optional	List of properties to compute (default: all)

Request

POST /v1/molecules/analyze
{
  "smiles": "CC(=O)Oc1ccccc1C(=O)O",
  "properties": ["logP", "tpsa", "admet"]
}

POST /v1/molecules/find-similar — Find similar molecules

Parameter	Type	Required	Description
`smiles`	string	Required	Query SMILES string
`threshold`	float	Optional	Tanimoto similarity threshold (default: 0.7)
`limit`	integer	Optional	Max results to return (default: 50)

Request

POST /v1/molecules/find-similar
{
  "smiles": "CC(=O)Oc1ccccc1C(=O)O",
  "threshold": 0.8,
  "limit": 20
}

GET /v1/molecules/{id} — Get molecule by ID

Parameter	Type	Required	Description
`id`	string (path)	Required	Molecule ID (e.g. `mol_a1b2c3d4e5f6g7h8`)

Request

GET /v1/molecules/mol_a1b2c3d4e5f6g7h8

POST /v1/molecules/batch — Batch property prediction

Parameter	Type	Required	Description
`smiles_list`	array	Required	Array of SMILES strings (max 100)
`properties`	array	Optional	Properties to compute

Request

POST /v1/molecules/batch
{
  "smiles_list": [
    "CC(=O)Oc1ccccc1C(=O)O",
    "CC(C)Cc1ccc(cc1)C(C)C(=O)O"
  ]
}

Docking

POST /v1/docking/run — Run molecular docking

Parameter	Type	Required	Description
`protein_pdb`	string	Required	PDB ID or uploaded protein structure
`ligand_smiles`	string	Required	SMILES of ligand to dock
`center`	object	Optional	Binding site center {x, y, z}
`box_size`	object	Optional	Search box dimensions {x, y, z}
`exhaustiveness`	integer	Optional	Docking exhaustiveness (default: 8)

Request

POST /v1/docking/run
{
  "protein_pdb": "6LU7",
  "ligand_smiles": "CC(=O)Oc1ccccc1C(=O)O",
  "exhaustiveness": 16
}

GET /v1/docking/{job_id} — Get docking results

Parameter	Type	Required	Description
`job_id`	string (path)	Required	Docking job ID returned from submit

Response

{
  "job_id": "dock_x7y8z9",
  "status": "completed",
  "best_score": -8.4,
  "num_poses": 9,
  "runtime_seconds": 42
}

GET /v1/docking/{job_id}/poses — Get binding poses

Parameter	Type	Required	Description
`job_id`	string (path)	Required	Docking job ID
`format`	string (query)	Optional	Output format: `pdb`, `sdf`, `mol2` (default: pdb)

Request

GET /v1/docking/dock_x7y8z9/poses?format=sdf

QSAR

POST /v1/qsar/train — Train QSAR model

Parameter	Type	Required	Description
`training_data`	array	Required	Array of {smiles, activity} objects
`model_type`	string	Optional	`random_forest`, `gradient_boost`, `neural_net`
`fingerprint`	string	Optional	Fingerprint type: `morgan`, `maccs`, `rdkit`
`name`	string	Optional	Model name for future reference

Request

POST /v1/qsar/train
{
  "training_data": [
    {"smiles": "CCO", "activity": 5.2},
    {"smiles": "CCCO", "activity": 4.8}
  ],
  "model_type": "gradient_boost",
  "fingerprint": "morgan",
  "name": "hERG_model_v1"
}

POST /v1/qsar/predict — Predict with trained model

Parameter	Type	Required	Description
`model_id`	string	Required	Trained model ID
`smiles_list`	array	Required	SMILES strings to predict

Request

POST /v1/qsar/predict
{
  "model_id": "qsar_m1n2o3",
  "smiles_list": ["c1ccccc1", "CC(=O)O"]
}

GET /v1/qsar/models — List your models

Parameter	Type	Required	Description
`page`	integer (query)	Optional	Page number (default: 1)
`limit`	integer (query)	Optional	Results per page (default: 20)

Request

GET /v1/qsar/models?page=1&limit=10

Molecular Dynamics

POST /v1/dynamics/run — Run MD simulation

Parameter	Type	Required	Description
`protein_pdb`	string	Required	PDB ID or structure data
`ligand_smiles`	string	Optional	SMILES of bound ligand
`duration_ns`	float	Optional	Simulation duration in nanoseconds (default: 10)
`force_field`	string	Optional	`amber99sb`, `charmm36`, `opls-aa`
`temperature`	float	Optional	Temperature in Kelvin (default: 300)

Request

POST /v1/dynamics/run
{
  "protein_pdb": "6LU7",
  "ligand_smiles": "CC(=O)Oc1ccccc1C(=O)O",
  "duration_ns": 10,
  "force_field": "amber99sb"
}

GET /v1/dynamics/{job_id} — Get simulation status

Parameter	Type	Required	Description
`job_id`	string (path)	Required	MD simulation job ID

Response

{
  "job_id": "md_p1q2r3",
  "status": "running",
  "progress": 0.45,
  "elapsed_seconds": 3420,
  "estimated_remaining": 4180
}

GET /v1/dynamics/{job_id}/trajectory — Get trajectory data

Parameter	Type	Required	Description
`job_id`	string (path)	Required	MD simulation job ID
`format`	string (query)	Optional	`xtc`, `trr`, `pdb` (default: xtc)
`frames`	string (query)	Optional	Frame range, e.g. `0-100`

Request

GET /v1/dynamics/md_p1q2r3/trajectory?format=xtc&frames=0-100

RL Training

POST /v1/rl/train — Train RL model for molecule generation

Parameter	Type	Required	Description
`scaffold`	string	Optional	Scaffold SMILES to build upon
`objectives`	object	Required	Optimization objectives (e.g. logP range, QED)
`num_molecules`	integer	Optional	Number of candidates to generate (default: 100)
`diversity_filter`	float	Optional	Minimum Tanimoto distance between results (default: 0.3)

Request

POST /v1/rl/train
{
  "scaffold": "c1ccc2[nH]ccc2c1",
  "objectives": {
    "logP": [1.0, 3.0],
    "qed": {"min": 0.6}
  },
  "num_molecules": 100
}

GET /v1/generate/{job_id} — Get generation results

Parameter	Type	Required	Description
`job_id`	string (path)	Required	Generation job ID

Response

{
  "job_id": "gen_s4t5u6",
  "status": "completed",
  "molecules": [
    {"smiles": "c1ccc2c(c1)cc(n2)CCN", "qed": 0.72, "logP": 1.85},
    {"smiles": "c1ccc2c(c1)cc(n2)CCCO", "qed": 0.68, "logP": 2.11}
  ],
  "total_generated": 100
}

FEP

POST /v1/fep/run — Run FEP calculation

Parameter	Type	Required	Description
`protein_pdb`	string	Required	Protein structure (PDB ID or data)
`ligand_a`	string	Required	SMILES of reference ligand
`ligand_b`	string	Required	SMILES of target ligand
`num_lambda`	integer	Optional	Number of lambda windows (default: 12)

Request

POST /v1/fep/run
{
  "protein_pdb": "6LU7",
  "ligand_a": "CC(=O)Oc1ccccc1C(=O)O",
  "ligand_b": "CC(C)Cc1ccc(cc1)C(C)C(=O)O",
  "num_lambda": 16
}

GET /v1/fep/{job_id} — Get FEP results

Parameter	Type	Required	Description
`job_id`	string (path)	Required	FEP job ID

Response

{
  "job_id": "fep_v7w8x9",
  "status": "completed",
  "ddG": -2.34,
  "ddG_error": 0.42,
  "convergence": "good"
}

Knowledge Graph

GET /v1/graph/molecules — Query molecule graph

Parameter	Type	Required	Description
`query`	string (query)	Optional	Search query (name, SMILES, or property filter)
`depth`	integer (query)	Optional	Traversal depth (default: 1, max: 3)
`limit`	integer (query)	Optional	Max results (default: 50)

Request

GET /v1/graph/molecules?query=aspirin&depth=2&limit=25

POST /v1/graph/relationships — Add relationship

Parameter	Type	Required	Description
`from_id`	string	Required	Source node ID
`to_id`	string	Required	Target node ID
`type`	string	Required	Relationship type (e.g. `similar_to`, `targets`)
`properties`	object	Optional	Edge metadata

Request

POST /v1/graph/relationships
{
  "from_id": "mol_a1b2c3d4e5f6g7h8",
  "to_id": "mol_i9j0k1l2m3n4o5p6",
  "type": "similar_to",
  "properties": {"tanimoto": 0.85}
}

GET /v1/graph/export — Export graph data

Parameter	Type	Required	Description
`format`	string (query)	Optional	`json`, `csv`, `graphml` (default: json)
`include_edges`	boolean (query)	Optional	Include edges in export (default: true)

Request

GET /v1/graph/export?format=graphml&include_edges=true

Tasks

GET /v1/tasks/{task_id} — Get task status

Parameter	Type	Required	Description
`task_id`	string (path)	Required	Task ID from any async operation

Response

{
  "task_id": "task_abc123",
  "type": "docking",
  "status": "completed",
  "created_at": "2026-03-05T10:30:00Z",
  "completed_at": "2026-03-05T10:31:42Z",
  "result_url": "/v1/docking/dock_x7y8z9"
}

GET /v1/tasks — List your tasks

Parameter	Type	Required	Description
`status`	string (query)	Optional	Filter: `pending`, `running`, `completed`, `failed`
`type`	string (query)	Optional	Filter by task type
`page`	integer (query)	Optional	Page number (default: 1)

Request

GET /v1/tasks?status=running&type=docking&page=1

DELETE /v1/tasks/{task_id} — Cancel task

Parameter	Type	Required	Description
`task_id`	string (path)	Required	Task ID to cancel

Request

DELETE /v1/tasks/task_abc123

Response

{
  "task_id": "task_abc123",
  "status": "cancelled",
  "message": "Task cancelled successfully"
}

IP Radar (Patent Landscaping)

POST /v1/patents/search — Search & analyze patent landscape

Searches global patent databases and uses MolForge AI to categorize claims, identify white spaces, and generate an IP landscape report.

Parameter	Type	Required	Description
`query`	string	Required	Drug name, CAS number, or SMILES
`max_patents`	integer	Optional	Max patents to analyze (default: 25, max: 100)

Request

POST /v1/patents/search
Content-Type: application/json
Authorization: Bearer YOUR_TOKEN

{
  "query": "Ticagrelor",
  "max_patents": 50
}

Response

{
  "query": "Ticagrelor",
  "stats": { "total_patents": 23, "by_category": {...} },
  "patents": [{ "patent_id": "US20210123456A1", ... }],
  "white_spaces": [{ "domain": "Polymorphism", ... }],
  "executive_summary": "...",
  "freedom_to_operate_assessment": "..."
}

Credit Costs

Each API operation consumes credits based on computational complexity. Monitor your usage on the dashboard.

Operation	Credits	Example
Property Lookup	1	Single SMILES → ADMET
Similarity Search	5	Find top-50 similar
Docking Job	50	Single protein-ligand dock
QSAR Training	200	Train custom model
MD Simulation	100/ns	10ns simulation = 1,000 cr
RL Generation	1,000	Generate 100 candidates
FEP Calculation	500	Single FEP run
Graph Query	1	Single graph traversal
IP Radar Scan	25	Patent landscape for one molecule

Code Examples

Get started quickly with examples in your preferred language.

Python

import requests

API_KEY = "mf_your_api_key"
BASE_URL = "https://api.molforge.ai/v1"

# Get molecular properties
response = requests.post(
    f"{BASE_URL}/molecules/analyze",
    headers={"X-API-Key": API_KEY},
    json={"smiles": "CC(=O)Oc1ccccc1C(=O)O"}
)
print(response.json())

# Run a docking job
dock_response = requests.post(
    f"{BASE_URL}/docking/run",
    headers={"X-API-Key": API_KEY},
    json={
        "protein_pdb": "6LU7",
        "ligand_smiles": "CC(=O)Oc1ccccc1C(=O)O"
    }
)
job_id = dock_response.json()["job_id"]

# Poll for results
import time
while True:
    result = requests.get(
        f"{BASE_URL}/docking/{job_id}",
        headers={"X-API-Key": API_KEY}
    ).json()
    if result["status"] == "completed":
        print(f"Best score: {result['best_score']}")
        break
    time.sleep(5)

JavaScript

const API_KEY = 'mf_your_api_key';
const BASE_URL = 'https://api.molforge.ai/v1';

// Get molecular properties
const response = await fetch(`${BASE_URL}/molecules/analyze`, {
  method: 'POST',
  headers: {
    'X-API-Key': API_KEY,
    'Content-Type': 'application/json'
  },
  body: JSON.stringify({ smiles: 'CC(=O)Oc1ccccc1C(=O)O' })
});
const data = await response.json();
console.log(data);

// Run docking job
const dockRes = await fetch(`${BASE_URL}/docking/run`, {
  method: 'POST',
  headers: {
    'X-API-Key': API_KEY,
    'Content-Type': 'application/json'
  },
  body: JSON.stringify({
    protein_pdb: '6LU7',
    ligand_smiles: 'CC(=O)Oc1ccccc1C(=O)O'
  })
});
const { job_id } = await dockRes.json();

// Poll for results
const pollResult = async () => {
  while (true) {
    const res = await fetch(`${BASE_URL}/docking/${job_id}`, {
      headers: { 'X-API-Key': API_KEY }
    });
    const result = await res.json();
    if (result.status === 'completed') return result;
    await new Promise(r => setTimeout(r, 5000));
  }
};
const result = await pollResult();

curl

# Get molecular properties
curl -X POST https://api.molforge.ai/v1/molecules/analyze \
  -H "X-API-Key: mf_your_api_key" \
  -H "Content-Type: application/json" \
  -d '{"smiles": "CC(=O)Oc1ccccc1C(=O)O"}'

# Submit a docking job
curl -X POST https://api.molforge.ai/v1/docking/run \
  -H "X-API-Key: mf_your_api_key" \
  -H "Content-Type: application/json" \
  -d '{"protein_pdb": "6LU7", "ligand_smiles": "CC(=O)Oc1ccccc1C(=O)O"}'

# Get docking results
curl https://api.molforge.ai/v1/docking/dock_x7y8z9 \
  -H "X-API-Key: mf_your_api_key"

# List your tasks
curl https://api.molforge.ai/v1/tasks?status=running \
  -H "X-API-Key: mf_your_api_key"

# Cancel a task
curl -X DELETE https://api.molforge.ai/v1/tasks/task_abc123 \
  -H "X-API-Key: mf_your_api_key"

Computational Methodology

MolForge integrates established open-source tools for molecular computations.

Property calculations use RDKit (open-source cheminformatics). Docking uses AutoDock Vina. MD simulations use OpenMM/GROMACS. ADMET predictions use machine learning models trained on public datasets. See individual tool documentation for algorithmic details and references.

Ready to Build with MolForge?

Start integrating molecular intelligence into your applications today.

Get API Key View Pricing

API Documentation

Foundation modelsfor drug discovery.

Quick Start

Get Your API Key

Make Your First Request

Authentication

API Key Header

Bearer Token

PDF Reports

Rate Limits

API Endpoints

Molecules

Docking

QSAR

Molecular Dynamics

RL Training

FEP

Knowledge Graph

Tasks

IP Radar (Patent Landscaping)

Credit Costs

Code Examples

Computational Methodology

Ready to Build with MolForge?

Foundation models
for drug discovery.