openfold3

name	openfold3
description	Structure prediction using OpenFold3, an open-weights PyTorch reproduction of AlphaFold3 from the AlQuraishi Lab. Use this skill when predicting protein/nucleic-acid/ligand complex structures with an Apache-2.0-licensed AF3 reimplementation.
license	Apache-2.0
origin	openai4s
category	biomodels
requirements	["gpu"]
capabilities	{"network":{"mode":"raw_required","domains":[]}}
metadata	{"display-name":"OpenFold3","third_party":[{"kind":"weights","name":"OpenFold3","provider":"OpenFold Consortium","license":"Apache-2.0","terms_url":"https://github.com/aqlaboratory/openfold-3/blob/main/LICENSE"},{"kind":"service","name":"ColabFold MSA server (api.colabfold.com)","provider":"Steinegger Lab","info_url":"https://github.com/sokrypton/ColabFold/wiki"}]}

OpenFold3 Structure Prediction

Prerequisites

Requirement	Minimum	Recommended
Python	3.10+	3.11
CUDA	12.1+	12.4+
GPU VRAM	24GB	80GB (H100)
RAM	32GB	64GB
Disk (weights)	3GB	-

How to run

Installation

pip install 'openfold3[cuequivariance]==0.4.1'

The default attention kernel is DeepSpeed DS4Sci_EvoformerAttention. If DeepSpeed is unavailable, switch to the cuEquivariance triangle kernels (no build-from-source) by overriding the eval memory settings in model_config.py (use_deepspeed_evo_attention: False, use_cueq_triangle_kernels: True). Some pre-built environments already ship this override; check before re-patching.

Weights

Apache-2.0, ~2.3 GB from HF OpenFold/OpenFold3. The repo is gated (auto-approval) — accept the access form on the HF model page and authenticate (huggingface-cli login or HF_TOKEN) before downloading:

export OPENFOLD_CACHE=~/.openfold3
huggingface-cli download OpenFold/OpenFold3 checkpoints/of3-p2-155k.pt \
  --local-dir "$OPENFOLD_CACHE"

run_openfold will also auto-download to $OPENFOLD_CACHE on first run if egress is open and HF credentials are available (either HF_TOKEN or a prior huggingface-cli login) with repo access granted. The interactive setup_openfold helper exists but prompts on stdin; prefer the explicit download above for non-interactive runs.

Running

export OPENFOLD_CACHE=/path/to/cache
run_openfold predict \
  --query_json=queries.json \
  --output-dir out/ \
  --use-msa-server false \
  --use-templates false

run_openfold discovers the checkpoint under automatically. Only pass if you have a non-standard layout or multiple checkpoints and need to pin one explicitly.