Skip to content

Model Labs

Biohub

Publisher of the ESMFold-2, ESMC, and ESM3 open protein model lines.

Multimodal generationProtein designProtein embeddingsStructure predictionVariant effects13 models

Models on Rafflesia

ESMFold-2 Fast

activeMIT
biohub/esmfold/2-fast/fold

Predict one small protein structure directly from sequence with the frozen ESMC 6B backbone and the inference-optimized, non-MSA ESMFold-2 Fast checkpoint.

Structure predictionprotein_sequencestructure_artifact + confidence_measurements + serving_measurements
$0.01 / prediction

ESMFold-2

activeMIT
biohub/esmfold/2/fold

Predict protein monomer and complex structures directly from sequence or an optional MSA with ESMFold-2.

Structure predictionprotein_sequencestructure_artifact + confidence_measurements
$0.04 / prediction

ESMC 300M

activeMIT
biohub/esmc/300m/embed

Create pooled or residue-level protein sequence representations with the compact ESM Cambrian model.

Protein embeddingsprotein_sequenceembedding_vector
$0.0005 / 1K residues

ESMC 300M

activeMIT
biohub/esmc/300m/variant-effects

Score amino acid substitutions as ESM Cambrian masked marginals with the compact 300M checkpoint.

Variant effectsprotein_sequence + amino_acid_substitutionslog_likelihood_ratio + mean_log_probability
$0.0005 / 1K residues

ESMC 600M

activeMIT
biohub/esmc/600m/embed

Create higher-capacity pooled or residue-level protein sequence representations with ESM Cambrian.

Protein embeddingsprotein_sequenceembedding_vector
$0.001 / 1K residues

ESMC 600M

activeMIT
biohub/esmc/600m/variant-effects

Score amino acid substitutions as ESM Cambrian masked marginals with the 600M checkpoint.

Variant effectsprotein_sequence + amino_acid_substitutionslog_likelihood_ratio + mean_log_probability
$0.001 / 1K residues

ESMC 6B

activeMIT
biohub/esmc/6b/embed

Create pooled or residue-level protein representations with the 6B-parameter ESM Cambrian checkpoint used as the frozen language-model backbone of ESMFold-2.

Protein embeddingsprotein_sequenceembedding_vector
$0.008 / 1K residues

ESMC 6B

activeMIT
biohub/esmc/6b/variant-effects

Score amino acid substitutions as ESM Cambrian masked marginals with the 6B-parameter checkpoint.

Variant effectsprotein_sequence + amino_acid_substitutionslog_likelihood_ratio + mean_log_probability
$0.008 / 1K residues

ESM3 1.4B Open

activeMIT
biohub/esm3/1.4b-open/embed

Create pooled or residue-level protein representations with the open 1.4B-parameter ESM3 checkpoint.

Protein embeddingsprotein_sequenceembedding_vector
$0.002 / 1K residues

ESM3 1.4B Open

activeMIT
biohub/esm3/1.4b-open/fold

Predict a protein structure by decoding the ESM3 structure track from sequence with the open 1.4B-parameter checkpoint.

Structure predictionprotein_sequencestructure_artifact + confidence_measurements
$0.03 / prediction

ESM3 1.4B Open

activeMIT
biohub/esm3/1.4b-open/inverse-fold

Sample protein sequences conditioned on a fixed backbone by decoding the ESM3 sequence track with the open 1.4B-parameter checkpoint.

Protein designstructureprotein_sequence
$0.002 / designed sequence

ESM3 1.4B Open

activeMIT
biohub/esm3/1.4b-open/generate

Complete a partially specified protein across the sequence, structure, secondary structure, solvent accessibility, and function tracks with the open 1.4B-parameter ESM3 checkpoint.

Multimodal generationprotein_sequence + structure + function_textprotein_sequence + structure_artifact + confidence_measurements
$0.03 / sample

ESM3 1.4B Open

activeMIT
biohub/esm3/1.4b-open/variant-effects

Score amino acid substitutions with ESM3 masked marginals from the open 1.4B-parameter checkpoint.

Variant effectsprotein_sequence + amino_acid_substitutionslog_likelihood_ratio + mean_log_probability
$0.002 / 1K residues

About

Biohub publishes open-weight protein models in the ESM lineage. Rafflesia identifies callable variants by their upstream names: ESMFold-2 and ESMFold-2 Fast, ESMC at 300M, 600M, and 6B parameters, and the open 1.4B ESM3 checkpoint.

Each model is called through a stable provider/family/variant endpoint id under the runs API.

Use these models

Every endpoint speaks the same run lifecycle: POST the endpoint id and a typed input, get a run back with outputs, provenance, and measured billing.

run.sh
# one run lifecycle, every endpoint
curl https://api.rafflesia.ai/v1/runs \
  -H "Authorization: Bearer $RAFFLESIA_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{
        "endpoint": "biohub/esmfold/2-fast/fold",
        "input": { "sequence": "MKTAYIAK..." }
      }'