Boltz-2

Inputs and outputs

• Your files are mounted read-only under /inputs inside the container.
• Prediction outputs and metadata should be written to /outputs.
• A shared reference volume is mounted at /ref; Boltz-2 uses /ref/cache for model data (enforced automatically).
• The container entrypoint is boltz predict, so the job arguments are passed directly to that CLI.
• If your YAML does not include MSAs, set msa: empty for each protein chain, or Boltz-2 may fail expecting MSA input.

Example 1 — basic MSA-backed run

Use the default pattern shown on the New Job form: point Boltz-2 at your inputs directory, enable the hosted MSA server, and turn on potentials.

/inputs
--use_msa_server
--use_potentials
--out_dir=/outputs

• Upstream recommends YAML inputs; FASTA inputs are supported but deprecated upstream.
• Place your YAML/FASTA and any required assets under /inputs when uploading.
• --use_msa_server tells Boltz-2 to fetch MSAs via the configured remote server.
• --use_potentials enables the learned potentials during inference.
• --out_dir=/outputs directs all prediction artifacts to the job output volume.

Example 2 — local configuration file

You can also drive Boltz-2 with an explicit configuration file that lives under /inputs, while still using /ref/cache for model data.

/inputs/my_config.yaml
--out_dir=/outputs/my_config_run
--use_msa_server
--use_potentials

• /inputs/my_config.yaml is a Boltz configuration file you prepare based on the upstream examples.
• Outputs for this run will be written under /outputs/my_config_run.
• The platform automatically appends --cache=/ref/cache to use the shared cache path.

Minimal no-MSA YAML pattern:

sequences:
  - protein:
      id: A
      sequence: "MKTAYIAKQRQISFVKSHFSRQDILDLI"
      msa: empty

Example 3 — multimer (heterodimer) from YAML

Create a YAML input under /inputs/heterodimer.yaml with multiple protein chains:

sequences:
  - protein:
      id: A
      sequence: "MKTAYIAKQRQISFVKSHFSRQDILDLI"
      msa: empty
  - protein:
      id: B
      sequence: "GSHSMRYFYTAMSRPGRGEPRFIAVGYV"
      msa: empty

Then run:

/inputs/heterodimer.yaml
--use_msa_server
--use_potentials
--out_dir=/outputs/heterodimer

All CLI options

SubSeq passes arguments directly to boltz predict, except that any cache flag is normalized to --cache=/ref/cache.

--out_dir PATH
--cache PATH
--checkpoint PATH
--affinity_checkpoint PATH
--devices INT
--accelerator [gpu|cpu|tpu]
--recycling_steps INT
--sampling_steps INT
--diffusion_samples INT
--max_parallel_samples INT
--step_scale FLOAT
--output_format [pdb|mmcif]
--num_workers INT
--preprocessing-threads INT
--override
--seed INT
--use_msa_server
--msa_server_url STRING
--msa_pairing_strategy STRING
--msa_server_username STRING
--msa_server_password STRING
--api_key_header STRING
--api_key_value STRING
--max_msa_seqs INT
--subsample_msa
--num_subsampled_msa INT
--no_kernels
--use_potentials
--method STRING
--model [boltz1|boltz2]
--affinity_mw_correction
--sampling_steps_affinity INT
--diffusion_samples_affinity INT
--write_full_pae
--write_full_pde
--write_embeddings

For detailed defaults and the YAML schema, see the upstream Boltz prediction instructions.

Notes

• Boltz-2 jobs currently require a GPU; they are scheduled onto the GPU cluster with 1 GPU, 8 vCPU, and 32 GiB RAM as a baseline.
• For full CLI options and configuration schemas, see the Boltz GitHub repository (boltz-2 section).
• To start a job from the UI, go to New Job → Boltz-2 and paste one of the argument blocks above.