Fastfold provides managed structure prediction, workflow execution, file-backed pipelines, and agent integrations through one platform.

Why use Fastfold

• Scientists + agents: collaborate with AI agents to plan, run, and iterate on folding experiments.
• Managed infrastructure: run jobs and workflows on managed compute without handling DevOps yourself.
• Reproducibility first: keep runs versioned and traceable so results are easy to compare and share.
• Flexible integrations: start with the Python SDK and CLI, drop to raw REST when you need full control, or work through agents.

Choose your path

• Python SDK and CLI for notebooks, scripts, CI, payload files, YAML submit, and workflow helpers: /sdk/install
• Agents for Agent CLI, skills, and PyMOL Agent: /agents
• REST API for direct HTTP integrations and endpoint reference: /api

Python SDK and CLI

Use the SDK when you want typed helpers for fold jobs, workflow helpers for OpenMM, OpenMMDL, and BoltzGen, plus access to raw payload workflows when needed.

pip install fastfold-ai

from fastfold import Client

client = Client()

job = client.fold.create(
    sequence="LLGDFFRKSKEKIGKEFKRIVQRIKDFLRNLVPRTES",
    model="boltz-2",
    is_public=True,
)

results = client.jobs.wait_for_completion(job.id, poll_interval=5.0, timeout=900.0)
print("Status:", results.job.status)
print("CIF URL:", results.cif_url())
print("Mean PLDDT:", results.metrics().mean_PLDDT)
print("Viewer:", results.get_viewer_link())

Use the SDK CLI (fastfold-cli) when you want the same platform features in shell scripts or CI:

fastfold-cli fold --sequence "LLGDFFRKSKEKIGKEFKRIVQRIKDFLRNLVPRTES" --model boltz-2
fastfold-cli jobs from-yaml --file fastfold/examples/fold/boltz2_affinity_input.yaml --model boltz-2
fastfold-cli workflows openmm from-manual-files --pdb ./protein.pdb --pae ./pae.json --simulation-name AF-P00698 --sim-length-ns 10 --box-length 50

The SDK section under /sdk includes install, overview, SDK CLI recipes, fold jobs, OpenMM, OpenMMDL, BoltzGen, generic workflows, webhooks, and report sharing.

Agent CLI

Use the Agent CLI when you want Fastfold to plan and run multi-step research flows from natural-language prompts.

uv tool install "fastfold-agent-cli[all]" --python 3.10

fastfold setup
fastfold "Fold this sequence with boltz-2 and show me the binding pockets: MALWMRLLPLL..."

Install, setup, and command reference → Agent CLI

Skills

If you use coding agents, install the official Fastfold skills package to expose fold and workflow operations in a reusable agent-friendly format.

npx skills add fastfold-ai/skills

• Skills repo: github.com/fastfold-ai/skills
• Skills listing: skills.sh/fastfold-ai/skills
• Detailed guide: /agents/skills

PyMOL Agent

Bring Fastfold skills directly into PyMOL with a chat-driven plugin workflow.

curl -LsSf http://fastfold.ai/pymol-agent/install.sh | sh

Install, setup, and usage guide → PyMOL Agent

llms.txt

This docs site includes AI-friendly exports:

• /llms.txt: concise index with documentation links.
• /llms-full.txt: full combined markdown content.

You can also append .mdx to docs paths to retrieve markdown content for a specific page (for example, /sdk/install.mdx).

Next steps

Start with the integration path that matches how you want to work:

• First SDK or CLI run: /quick-start
• Install and configure the SDK: /sdk/install
• Browse SDK recipes by task: /sdk
• Read the REST reference: /api