Access state-of-the-art models through clean, well-documented endpoints. No infrastructure to manage.
Predict 3D protein structures from amino acid sequences using ESMFold. Get PDB files and confidence scores.
Calculate molecular properties, convert between SMILES/InChI formats, and compute Tanimoto similarity.
Dock small molecules to protein targets with AutoDock Vina. Get binding affinities and 3D poses.
Predict absorption, distribution, metabolism, excretion, and toxicity properties for drug candidates.
Analyze variant effects, predict pathogenicity scores, and explore gene-disease associations.
End-to-end neoantigen vaccine pipeline: mutation analysis, MHC binding, mRNA construct design.
40+ tools via Model Context Protocol. Let Claude, GPT, or custom agents call science APIs natively.
One Python package. Sync and async clients. Type-safe responses. pip install scirouter.
No infrastructure to set up. No GPUs to manage. No Docker images to build.
Sign up in 30 seconds. No credit card needed. Your key is ready immediately with 500 free credits per month.
pip install scirouter — one package for every model. Or use the REST API directly with curl, fetch, or any HTTP client.
Fold proteins, analyze molecules, dock drugs, design vaccines. Same auth, same format, same reliability across all endpoints.
Whether you're a solo researcher, a pharma team, or an AI agent — SciRouter fits your workflow.
Screen compounds, predict ADMET properties, dock molecules to targets, and design mRNA vaccines — all through one API. Replace five vendor contracts with a single integration.
40+ MCP tools let Claude, GPT, and custom agents call real science models. Build autonomous research assistants that fold proteins and analyze molecules in-context.
Focus on science, not infrastructure. Get started in minutes with the Python SDK or REST API. Free tier included — no grants required to prototype.
Paste a protein sequence or molecule SMILES below and get instant results. This is exactly what the API returns.
client.proteins.fold("MALWMRLLPLLALLALWGPD...")Click “Predict Structure” to see results
Install the SDK, create a client, and start calling models. It really is that simple.
from scirouter import SciRouter
client = SciRouter(api_key="sk-sci-...")
# Predict protein structure
job = client.proteins.fold(
"MVLSPADKTNVKAAWGKVGAHAGEYGAEALERMFL"
)
print(job.status) # "completed"
print(job.result.pdb) # Full PDB structure
# Molecular properties
props = client.chemistry.properties("CCO")
print(props.molecular_weight) # 46.07
print(props.logp) # -0.31REST, SDK, or MCP — integrate however you build. One API key, every access method.
40+ tools for Claude, GPT, and custom AI agents via Model Context Protocol
Type-safe sync & async clients. pip install scirouter and start building.
Standard JSON REST endpoints. Works with any language, framework, or platform.
Tool schemas in OpenAI function format. Drop-in compatible with agent frameworks.
Start free. Upgrade when you need more. No surprises.
500 credits/month (750 for academic)
Get started with science APIs at no cost.
5,000 credits/month
For researchers who need more compute and faster results.
New tools, API features, research guides, and model benchmarks — delivered monthly. No spam.