ESMFold

Predict a protein structure with the ESM2 language model. Start from a single sequence without an MSA of related sequences.

API/CLI tool ID: esmfold

Understanding inputs and outputs

ESMFold predicts a protein’s three-dimensional structure from its amino acid sequence. Each amino acid in the chain is called a residue. It does not require an input MSA file containing an alignment of related sequences.

pLDDT measures the model’s confidence in local structure around each residue; the results table shows an average. Some regions may have low confidence even when the average is high. pTM estimates confidence in the overall arrangement of the structure. Higher values indicate greater model confidence, not experimental accuracy or the probability of successful binding.

The pLDDT in the result table and in plddt.csv is on a 0–100 scale. The B-factor column of predicted_structure.pdb holds the same quantity as a 0–1 fraction. Multiply by 100 before passing that structure file to another tool or viewer that reads B-factor as 0–100 pLDDT. AlphaFold and ColabFold result files are written on the 0–100 scale from the start, so the scales differ.

Input settings

SettingFormatRequiredDefaultDescription and constraints
sequencesequenceRequired—Characters ACDEFGHIKLMNPQRSTVWY: , Unknown residue X-stripped, Example "MTYKLILNGKTLKGETTTEAVDAATAEKVFKQYANDNGVDGEWTYDDATKTFTVTE"

Output metrics

ColumnDisplay nameRangeDescription
plddtpLDDT0~100Mean per-residue pLDDT
ptmpTM0~1—

See Reading results for metric definitions and interpretation. A range shows possible values; whether higher or lower is better depends on the metric.

Execution availability

Open an execution plan in your project to see available tool versions and costs. Review and approve the inputs, settings, and terms before running.

License

Code, model weights, and accompanying data have their own terms. The table summarizes terms from official sources; check your execution plan for the configuration you will use. License texts are available in Third-party notices.

ComponentLicenseCommercial useSourceNotes
CodeMITAllowed, subject to notice requirementsLicense text—
Model weightsMITAllowed, subject to notice requirementsWeight distribution metadataApplies to the model weights distributed as facebook/esmfold_v1 on Hugging Face.

The official source of the weights used here is facebook/esmfold_v1 on Hugging Face.

Checked on 2026-10-24.

References

On this page