ChEMBL Drug & Molecule Scraper — Free Chemistry Data, No API Key

Scrape drugs and bioactive molecules from ChEMBL (EMBL-EBI) into clean, structured JSON, CSV or Excel — no API key, no login, no proxy required. Search by drug name or browse the entire ChEMBL molecule table, and get molecular formula, molecular weight, clinical phase, Lipinski properties, SMILES and InChI keys in one row per molecule.

ChEMBL is the reference open database for medicinal chemistry: 2.9M+ molecules, curated bioactivities and approved-drug annotations. This actor turns its web services into a ready-to-use dataset for drug discovery, cheminformatics and pharma research pipelines.

What it does

• Search mode — full-text ChEMBL search for any drug or molecule name (aspirin, imatinib, semaglutide, or a ChEMBL ID like CHEMBL25).
• Browse mode — walk the complete ChEMBL molecule table page by page to build bulk datasets.
• One dataset item per molecule, deduplicated by ChEMBL ID across all queries.
• Every numeric field is coerced to a real number, so molecular_weight, alogp and psa are ready for pandas / RDKit without cleanup.

Input

Field	Type	Default	Description
mode	select (search / browse)	search	Search by query, or browse all molecules
queries	array of strings	["aspirin", "imatinib"]	Drug/molecule names to search (search mode only)
maxItems	integer	100 (max 1000)	Maximum molecules saved to the dataset

{
  "mode": "search",
  "queries": ["aspirin", "imatinib", "atorvastatin"],
  "maxItems": 300
}

Output

{
  "chembl_id": "CHEMBL25",
  "name": "ASPIRIN",
  "max_phase": 4.0,
  "molecule_type": "Small molecule",
  "first_approval": 1950,
  "oral": true,
  "parenteral": false,
  "topical": false,
  "black_box_warning": false,
  "withdrawn": false,
  "molecular_weight": 180.16,
  "alogp": 1.31,
  "hba": 3,
  "hbd": 1,
  "psa": 63.6,
  "rotatable_bonds": 2,
  "ro5_violations": 0,
  "formula": "C9H8O4",
  "smiles": "CC(=O)Oc1ccccc1C(=O)O",
  "inchi_key": "BSYNRYMUTXBXSQ-UHFFFAOYSA-N",
  "url": "https://www.ebi.ac.uk/chembl/compound_report_card/CHEMBL25",
  "source": "chembl",
  "scraped_at": "2026-07-28T11:36:59+00:00"
}

All fields are nullable — ChEMBL does not compute properties for every entry (biotherapeutics, unknowns), so missing values come back as null instead of breaking your pipeline.

Pricing

Runs on the pay-per-usage model — roughly $0.5 per 1,000 molecules at 512 MB memory. A typical 100-molecule search run costs a few cents and finishes in well under a minute.

Use cases

• Drug discovery — pull approved-drug sets by clinical phase, first approval year and route of administration to seed target or repurposing screens.
• Cheminformatics — export SMILES and InChI keys in bulk for RDKit fingerprints, similarity search, clustering and QSAR modelling.
• Pharma research — track black-box warnings, withdrawn drugs and Lipinski rule-of-five violations across a compound portfolio.
• Datasets & ML — build training tables of molecular descriptors (MW, AlogP, HBA, HBD, PSA, rotatable bonds) for property-prediction models.

Tips

• Use browse mode with maxItems: 1000 to sample the database, then search mode for targeted follow-ups.
• Search matches synonyms and trade names too — aspirin returns aspirin salts, esters and combinations, all with their own ChEMBL IDs.
• Join results on inchi_key to merge ChEMBL data with PubChem, DrugBank or your internal compound registry.

Related actors

• PubChem Compounds Scraper — chemical structures and properties from PubChem
• UniProt Scraper — protein sequences, functions and targets
• openFDA Scraper — drug labels, adverse events and recalls
• Clinical Trials Scraper — study records from ClinicalTrials.gov

Legal & data source

Data comes from the public ChEMBL web services operated by EMBL-EBI. ChEMBL data is released under CC BY-SA 3.0 — attribute ChEMBL when you republish. Only publicly accessible, non-personal scientific data is collected. Please keep run volumes reasonable to respect EMBL-EBI's shared infrastructure.