Skip to content
@BioLM

BioLM

BioLM

BioLM — hosted platform or open source deployments, one SDK

Use the hosted BioLM platform at biolm.ai or deploy models yourself with biolm-hub. Either way, you use the same Python SDK and CLI: biolm-sdk.

Website Docs PyPI Slack Feedback


Open source + commercial platform (how it fits together)

BioLM is a commercial hosted platform and an open-source software ecosystem.

  • Hosted platform (commercial): managed infrastructure, hosted APIs, and platform features for running models and workflows on BioLM’s servers: biolm.ai
  • Open source (community): reusable tools and model deployments you can run in your own environment:
    • biolm-sdk — Python SDK + CLI
    • biolm-hub — deploy open models in your own Modal workspace

We love open source because it makes the ecosystem more transparent, more reproducible, and easier to build on — for humans and for agents. If you’re contributing models, docs, integrations, or ideas, you’re helping expand what the whole community can do.


Primary repos

Python SDK and biolm CLI for the hosted platform and biolm-hub gateways.

  • biolm() one-liners, Model, and async BioLMApiClient
  • CLI: biolm model, biolm protocol, biolm hub, workspaces & datasets
  • Optional pipeline framework (DuckDB caching, multi-stage DAGs)
  • Python 3.8+ · Apache-2.0
pip install biolm-sdk
biolm login
biolm model run esm2-8m encode -i seq.json

Documentation → · PyPI →

Standardized, agent-first catalog of open biological ML models. Deploy on Modal in a few commands.

  • ESM2, ESMFold, ProteinMPNN, DNABERT, AntiFold, and more
  • Uniform verbs: predict, fold, encode, generate, score, log_prob
  • Per-model knowledge graph (license, papers, when-to-use)
  • bh deploy + bh serve → local HTTP API + browser catalog UI
git clone https://github.com/BioLM/biolm-hub
cd biolm-hub && make install && source .venv/bin/activate
bh setup && bh deploy esm2 && bh serve

Philosophy → · Contributing →


Use them together

Point the SDK at your running hub, keep platform auth and protocols on biolm.ai:

# Terminal 1 — in biolm-hub
bh serve

# Terminal 2 — anywhere
pip install biolm-sdk
biolm hub set
biolm model list
biolm model run esm2-8m encode -i sequences.json
from biolm import biolm

# Hosted platform (after biolm login or BIOLM_TOKEN)
result = biolm(entity="esmfold", action="predict", type="sequence", items=["MKTAYIAKQRQ"])

# Same API against a hub gateway (after biolm hub set)
result = biolm(entity="esm2-8m", action="encode", type="sequence", items="MSILVTRPSPAGEEL")

What you can build

Capability Examples
Embeddings & representations ESM2, ESMC, ESM-1v, DNABERT
Structure prediction & design ESMFold, ProteinMPNN, AntiFold, BoltzGen
Sequence generation & scoring ProGen2, DSM, ThermoMPNN
Multi-stage workflows Protocol pipelines via SDK; optional biolm.pipeline DAGs
Self-hosted inference Deploy hub models to your Modal workspace

Community & feedback

Slack biolm.ai/community — ask questions, share workflows
Roadmap & feature requests biolm.canny.io
SDK issues biolm-sdk/issues
Hub issues & model proposals biolm-hub/issues · discussions

Also in this org

Workflow integrations

Use BioLM from common pipeline and workflow frameworks:

Repo Framework
nf-biolm Nextflow
biolm-cwl Common Workflow Language (CWL)
biolm-galaxy Galaxy
biolm-kedro Kedro

JupyterLab

Repo Description
jupyterlab-biolm JupyterLab extension for BioLM
jupyterlab-mlflow JupyterLab extension for MLflow

Lab-in-the-Loop Transfer Protocol (LLTP)

Repo Description
lltp Lab-in-the-Loop Transfer Protocol
lltp-py Python reference SDK
lltp-js JavaScript/TypeScript reference SDK

New here? Install biolm-sdk, join us on Slack, or deploy your first model with biolm-hub.

Popular repositories Loading

  1. tutorials-and-guides tutorials-and-guides Public

    Notebooks demonstrating quick and powerful ways to use BioLM for bioinformatics and modeling tasks

    HTML 4 1

  2. nf-biolm nf-biolm Public

    Using BioLM within Nextflow pipelines.

    Python 4 1

  3. jupyterlab-mlflow jupyterlab-mlflow Public

    JupyterLab extension for MLFlow

    TypeScript 4

  4. biolm-sdk biolm-sdk Public

    Python SDK & CLI for BioLM.ai

    Python 3

  5. py-biolm py-biolm Public

    Python SDK for BioLM.ai API

    Python 1 1

  6. biolm-hub biolm-hub Public

    BioLM Hub — a standardized, agent-first catalog of open biological ML models that deploy on Modal.

    Python 1

Repositories

Showing 10 of 23 repositories

Top languages

Loading…

Most used topics

Loading…