Setting Up Your Environment for ESM3

Updated for 2026: Environment setup for ESM3 in 2026: Python isolation, Biohub SDK installation, secrets and test checks.

Create your environment

Start with a clean virtual environment. For example, on Linux or macOS:

python -m venv .venv
source .venv/bin/activate
python -m pip install --upgrade pip
python -m pip install esm

On Windows, activate the environment using the corresponding Scripts directory. Check the maintained repository for compatible Python and accelerator requirements.

Choose your inference path

Hosted Biohub inference requires an account, available API access and a secret token. Local inference requires a supported checkpoint and compatible hardware. These are different workflows; a library installation alone does not grant unlimited remote inference.

Keep secrets on the server

Store API credentials in the environment or secrets manager. Do not paste them into HTML, client-side JavaScript, public notebooks or screenshots. Establish usage alerts, spending limits and data-retention policies before automating requests.

Verify before scaling

Run the official quickstart with a small, non-sensitive input, then record model version and elapsed time. Test error handling and output shape before processing large batches. Do not assume old function names remain supported.

Evidence and further reading

Primary sources: Biohub ESM3 model, Biohub maintained ESM repository, and Biohub learning resources. Check model cards and release notes for version-specific instructions.

Editorial note: Updated October 2026. Distinguish scientific literature and current product documentation from speculative applications; verify current API details before running code.