Configuration

ESM-C

ESM C learns protein sequence representations for downstream analysis.

Sources (5)

evolutionaryscale/esm: README.md; evolutionaryscale/esm: _assets/ESM3_README.md; evolutionaryscale/esm: THIRD_PARTY_NOTICE.md; biohub/ESMC-6B: README.md; biohub/ESMC-6B: config.json · Official biohub/ESMC-6B card: Model Architecture, Parameters, Training Data, Training Procedure and Limitations; THIRD_PARTY_NOTICE.md

12 evaluations · 12 results

How it worksESMC workflow

Conceptual summary of the documented data flow; optional inputs and configured downstream stages must be reported for a reproducible evaluation.

Sources (5)

Overview

Model type

Protein representation transformer

Inputs

Protein amino-acid sequences.

Outputs

Final-layer or all-layer protein representations and masked-token outputs.

Access

Official project documentation and implementation: https://github.com/evolutionaryscale/esm

Sources (5)

limited source coverage · Automated source review, 2026-09-16. All specifications and missing details

Evaluations and results

12 evaluations · 12 results. Different protocols are not a single leaderboard.

Filter evaluations

Tested model or methodProtocolDatasetEvaluation setupMetricEvidence origin

Applied filters: All linked evaluations

Exact evaluated configurations and original reported results
Tested configuration	Protocol and dataset	Finding	Evidence and details
Configuration: ESM-C	Task: PFMBench ANTI-RES: Antibiotic resistance Dataset subset: Antibiotic resistance (PFMBench split)	0.673 accuracy fraction · higher Uncertainty: Not reported Coverage: Not reported scored / Not reported eligible	Author-reported evaluation · Source checked Methods, coverage and source ESM-C on PFMBench ANTI-RES: Antibiotic resistance Fine-tuned with an adapter under the PFMBench harness; train, validation and test counts are in Table 1. Aggregation: Not reported pfmbench primary benchmark evidence · Table 3, row(ESM-C), column(Anti.Res.)
Configuration: ESM-C	Task: PFMBench BINDING-DB: BindingDB Dataset subset: BindingDB (PFMBench split)	0 .20716 spearman correlation · higher Uncertainty: Not reported Coverage: Not reported scored / Not reported eligible	Author-reported evaluation · Source checked Methods, coverage and source ESM-C on PFMBench BINDING-DB: BindingDB Fine-tuned with an adapter under the PFMBench harness; train, validation and test counts are in Table 1. Aggregation: Not reported pfmbench primary benchmark evidence · Table 3, row(ESM-C), column(Bind. DB)
Configuration: ESM-C	Task: PFMBench CLONING-CLF: Cloning CLF Dataset subset: Cloning CLF (PFMBench split)	0.81 auroc fraction · higher Uncertainty: Not reported Coverage: Not reported scored / Not reported eligible	Author-reported evaluation · Source checked Methods, coverage and source ESM-C on PFMBench CLONING-CLF: Cloning CLF Fine-tuned with an adapter under the PFMBench harness; train, validation and test counts are in Table 1. Aggregation: Not reported pfmbench primary benchmark evidence · Table 3, row(ESM-C), column(Clo. CLF)
Configuration: ESM-C	Task: PFMBench DEEPLOC2: DeepLoc2 Multi Dataset subset: DeepLoc2 Multi (PFMBench split)	0.754 f1 fraction · higher Uncertainty: Not reported Coverage: Not reported scored / Not reported eligible	Author-reported evaluation · Source checked Methods, coverage and source ESM-C on PFMBench DEEPLOC2: DeepLoc2 Multi Fine-tuned with an adapter under the PFMBench harness; train, validation and test counts are in Table 1. Aggregation: Not reported pfmbench primary benchmark evidence · Table 3, row(ESM-C), column(DL2 M.)
Configuration: ESM-C	Task: PFMBench DEEPSOL: DeepSol Dataset subset: DeepSol (PFMBench split)	0.842 auroc fraction · higher Uncertainty: Not reported Coverage: Not reported scored / Not reported eligible	Author-reported evaluation · Source checked Methods, coverage and source ESM-C on PFMBench DEEPSOL: DeepSol Fine-tuned with an adapter under the PFMBench harness; train, validation and test counts are in Table 1. Aggregation: Not reported pfmbench primary benchmark evidence · Table 3, row(ESM-C), column(DeepSol)
Configuration: ESM-C	Task: PFMBench EC: Enzyme Commission Dataset subset: Enzyme Commission (PFMBench split)	0.717 f1 fraction · higher Uncertainty: Not reported Coverage: Not reported scored / Not reported eligible	Author-reported evaluation · Source checked Methods, coverage and source ESM-C on PFMBench EC: Enzyme Commission Fine-tuned with an adapter under the PFMBench harness; train, validation and test counts are in Table 1. Aggregation: Not reported pfmbench primary benchmark evidence · Table 3, row(ESM-C), column(EC)
Configuration: ESM-C	Task: PFMBench MAT-PROD: Material production Dataset subset: Material production (PFMBench split)	0.81 accuracy fraction · higher Uncertainty: Not reported Coverage: Not reported scored / Not reported eligible	Author-reported evaluation · Source checked Methods, coverage and source ESM-C on PFMBench MAT-PROD: Material production Fine-tuned with an adapter under the PFMBench harness; train, validation and test counts are in Table 1. Aggregation: Not reported pfmbench primary benchmark evidence · Table 3, row(ESM-C), column(Mat.Pro.)
Configuration: ESM-C	Task: PFMBench METAL-ION: Metal ion binding Dataset subset: Metal ion binding (PFMBench split)	0.702 accuracy fraction · higher Uncertainty: Not reported Coverage: Not reported scored / Not reported eligible	Author-reported evaluation · Source checked Methods, coverage and source ESM-C on PFMBench METAL-ION: Metal ion binding Fine-tuned with an adapter under the PFMBench harness; train, validation and test counts are in Table 1. Aggregation: Not reported pfmbench primary benchmark evidence · Table 3, row(ESM-C), column(M. I. Bin.)
Configuration: ESM-C	Task: PFMBench PDB-BIND: PDBbind Dataset subset: PDBbind (PFMBench split)	0.147 spearman correlation · higher Uncertainty: Not reported Coverage: Not reported scored / Not reported eligible	Author-reported evaluation · Source checked Methods, coverage and source ESM-C on PFMBench PDB-BIND: PDBbind Fine-tuned with an adapter under the PFMBench harness; train, validation and test counts are in Table 1. Aggregation: Not reported pfmbench primary benchmark evidence · Table 3, row(ESM-C), column(PDBBind)
Configuration: ESM-C	Task: PFMBench PROTEINGYM-ZS: ProteinGym zero-shot variant effect prediction Dataset subset: ProteinGym (PFMBench split)	0.434 spearman correlation · higher Uncertainty: Not reported Coverage: Not reported scored / Not reported eligible	Author-reported evaluation · Source checked Methods, coverage and source ESM-C on PFMBench PROTEINGYM-ZS: ProteinGym zero-shot variant effect prediction Scored zero-shot, with no fine-tuning, as reported in PFMBench Table 4. Aggregation: Not reported pfmbench primary benchmark evidence · Table 4, row(ESM-C), column(ProteinGym)
Configuration: ESM-C	Task: PFMBench SEC-STRUCT: Secondary structure Dataset subset: Secondary structure (PFMBench split)	0.768 accuracy fraction · higher Uncertainty: Not reported Coverage: Not reported scored / Not reported eligible	Author-reported evaluation · Source checked Methods, coverage and source ESM-C on PFMBench SEC-STRUCT: Secondary structure Fine-tuned with an adapter under the PFMBench harness; train, validation and test counts are in Table 1. Aggregation: Not reported pfmbench primary benchmark evidence · Table 3, row(ESM-C), column(Sec. Str.)
Configuration: ESM-C	Task: PFMBench STABILITY: TAPE_Stability Dataset subset: TAPE_Stability (PFMBench split)	0.3 spearman correlation · higher Uncertainty: Not reported Coverage: Not reported scored / Not reported eligible	Author-reported evaluation · Source checked Methods, coverage and source ESM-C on PFMBench STABILITY: TAPE_Stability Fine-tuned with an adapter under the PFMBench harness; train, validation and test counts are in Table 1. Aggregation: Not reported pfmbench primary benchmark evidence · Table 3, row(ESM-C), column(Stability)

Source checking is not independent reproduction. Release 2026-09-29-06401fd5b220.

Use this model

How it works, versions and access

Related profile: ESMC. This page retains the exact record and its evaluation context.

This configuration

Protein foundation model evaluated by the PFMBench authors under their fine-tuning protocol. Input family: Sequence.

record: ESM-C
configuration: Not reported
entity type: Configuration

How it works

ESM C learns protein sequence representations for downstream analysis. Pre-normalized transformer with rotary embeddings, SwiGLU feed-forward activations and no linear/layer-norm biases. The documented inputs are protein amino-acid sequences. The output consists of final-layer or all-layer protein representations and masked-token outputs.

Sources (5)

Versions and reproducibility

esmc-600m-2024-12 API identifier and biohub/ESMC-6B local example. The card specifies a2,048-token window after an initial 512-token training phase.

Sources (5)

Strengths, limitations and unresolved questions

Strengths and limitations

Strengths and considerations

The implementation supports local inference and a hosted interface, with downloadable model variants.
Sources (5)
evolutionaryscale/esm: README.md; evolutionaryscale/esm: _assets/ESM3_README.md; evolutionaryscale/esm: THIRD_PARTY_NOTICE.md; biohub/ESMC-6B: README.md; biohub/ESMC-6B: config.json · Official biohub/ESMC-6B card: Model Architecture, Parameters, Training Data, Training Procedure and Limitations; THIRD_PARTY_NOTICE.md

Limitations and conditions

Hosted API names and downloadable weights identify different versions. The main repository does not by itself provide a complete per-checkpoint training manifest.
Sources (5)
evolutionaryscale/esm: README.md; evolutionaryscale/esm: _assets/ESM3_README.md; evolutionaryscale/esm: THIRD_PARTY_NOTICE.md; biohub/ESMC-6B: README.md; biohub/ESMC-6B: config.json · Official biohub/ESMC-6B card: Model Architecture, Parameters, Training Data, Training Procedure and Limitations; THIRD_PARTY_NOTICE.md

Profile review details

Inspected pinned official documentation, relevant implementation files and named primary-paper sections. Claims are limited to those artifacts. Remaining field extraction and identity conflicts are explicit; no new performance claims, model runs or human review are implied.

Stable record: discovery-model-esmc

Specifications

Inputs, training, access and other details

Explanatory profile: limited source coverage · Automated source review, 2026-09-16. Review applies to the cited claims; unresolved fields are listed below. Numerical results retain their own review status.

Inputs, outputs and configuration
Property	Description and evidence
Model type	Protein representation transformer Sources (5) evolutionaryscale/esm: README.md; evolutionaryscale/esm: _assets/ESM3_README.md; evolutionaryscale/esm: THIRD_PARTY_NOTICE.md; biohub/ESMC-6B: README.md; biohub/ESMC-6B: config.json · Official biohub/ESMC-6B card: Model Architecture, Parameters, Training Data, Training Procedure and Limitations; THIRD_PARTY_NOTICE.md
Architecture	Pre-normalized transformer with rotary embeddings, SwiGLU feed-forward activations and no linear/layer-norm biases. Sources (5) evolutionaryscale/esm: README.md; evolutionaryscale/esm: _assets/ESM3_README.md; evolutionaryscale/esm: THIRD_PARTY_NOTICE.md; biohub/ESMC-6B: README.md; biohub/ESMC-6B: config.json · Official biohub/ESMC-6B card: Model Architecture, Parameters, Training Data, Training Procedure and Limitations; THIRD_PARTY_NOTICE.md
Inputs	Protein amino-acid sequences. Sources (5) evolutionaryscale/esm: README.md; evolutionaryscale/esm: _assets/ESM3_README.md; evolutionaryscale/esm: THIRD_PARTY_NOTICE.md; biohub/ESMC-6B: README.md; biohub/ESMC-6B: config.json · Official biohub/ESMC-6B card: Model Architecture, Parameters, Training Data, Training Procedure and Limitations; THIRD_PARTY_NOTICE.md
Outputs	Final-layer or all-layer protein representations and masked-token outputs. Sources (5) evolutionaryscale/esm: README.md; evolutionaryscale/esm: _assets/ESM3_README.md; evolutionaryscale/esm: THIRD_PARTY_NOTICE.md; biohub/ESMC-6B: README.md; biohub/ESMC-6B: config.json · Official biohub/ESMC-6B card: Model Architecture, Parameters, Training Data, Training Procedure and Limitations; THIRD_PARTY_NOTICE.md
Parameters	300M/30 layers, 600M/36 layers and 6B/80 layers in the checked model card. Sources (5) evolutionaryscale/esm: README.md; evolutionaryscale/esm: _assets/ESM3_README.md; evolutionaryscale/esm: THIRD_PARTY_NOTICE.md; biohub/ESMC-6B: README.md; biohub/ESMC-6B: config.json · Official biohub/ESMC-6B card: Model Architecture, Parameters, Training Data, Training Procedure and Limitations; THIRD_PARTY_NOTICE.md
Known versions	esmc-600m-2024-12 API identifier and biohub/ESMC-6B local example. Sources (5) evolutionaryscale/esm: README.md; evolutionaryscale/esm: _assets/ESM3_README.md; evolutionaryscale/esm: THIRD_PARTY_NOTICE.md; biohub/ESMC-6B: README.md; biohub/ESMC-6B: config.json · Official biohub/ESMC-6B card: Model Architecture, Parameters, Training Data, Training Procedure and Limitations; THIRD_PARTY_NOTICE.md
Training data	UniRef, MGnify and JGI protein sequences clustered at 70% identity. The card distinguishes83M, 372M and 2B clusters respectively from the number of repeated training tokens. Sources (5) evolutionaryscale/esm: README.md; evolutionaryscale/esm: _assets/ESM3_README.md; evolutionaryscale/esm: THIRD_PARTY_NOTICE.md; biohub/ESMC-6B: README.md; biohub/ESMC-6B: config.json · Official biohub/ESMC-6B card: Model Architecture, Parameters, Training Data, Training Procedure and Limitations; THIRD_PARTY_NOTICE.md
Training cutoff	The official ESMC-6B card identifies UniRef, MGnify and JGI clusters and training stages, but does not give a latest-sequence date shared across those corpora. · Not reported in inspected sources Sources (5) evolutionaryscale/esm: README.md; evolutionaryscale/esm: _assets/ESM3_README.md; evolutionaryscale/esm: THIRD_PARTY_NOTICE.md; biohub/ESMC-6B: README.md; biohub/ESMC-6B: config.json · Official biohub/ESMC-6B card: Model Architecture, Parameters, Training Data, Training Procedure and Limitations; THIRD_PARTY_NOTICE.md
Context limits	The card specifies a2,048-token window after an initial 512-token training phase. Sources (5) evolutionaryscale/esm: README.md; evolutionaryscale/esm: _assets/ESM3_README.md; evolutionaryscale/esm: THIRD_PARTY_NOTICE.md; biohub/ESMC-6B: README.md; biohub/ESMC-6B: config.json · Official biohub/ESMC-6B card: Model Architecture, Parameters, Training Data, Training Procedure and Limitations; THIRD_PARTY_NOTICE.md
Weights licence	MIT is declared alongside third-party notices in the checked6B card; THIRD_PARTY_NOTICE.md lists dependency licences. Sources (5) evolutionaryscale/esm: README.md; evolutionaryscale/esm: _assets/ESM3_README.md; evolutionaryscale/esm: THIRD_PARTY_NOTICE.md; biohub/ESMC-6B: README.md; biohub/ESMC-6B: config.json · Official biohub/ESMC-6B card: Model Architecture, Parameters, Training Data, Training Procedure and Limitations; THIRD_PARTY_NOTICE.md
Access	Official project documentation and implementation: https://github.com/evolutionaryscale/esm Sources (5) evolutionaryscale/esm: README.md; evolutionaryscale/esm: _assets/ESM3_README.md; evolutionaryscale/esm: THIRD_PARTY_NOTICE.md; biohub/ESMC-6B: README.md; biohub/ESMC-6B: config.json · Official biohub/ESMC-6B card: Model Architecture, Parameters, Training Data, Training Procedure and Limitations; THIRD_PARTY_NOTICE.md
Code licence	MIT Sources evolutionaryscale/esm: LICENSE.md · LICENSE.md: licence text

Evidence

Source checking verifies the cited claim or transcription. It does not establish independent reproduction.

Evidence table

Inspect claims, sources and review details

Trace each statement to its source and review. A context-only reference supports the record generally; it does not verify an individual field. Source checking does not reproduce an experiment.

One row per statement and cited source. Multiple citations are not independent evaluations. Shared locators are labelled explicitly.

1 evidence row matching the loaded filters

Claims, original sources and review scope · Release 2026-09-29-06401fd5b220
Property and statement	Original source and location	Review and provenance
Relationship: family discovery-model-esmc Individual claims	pfmbench primary benchmark evidence Original source ↗ Section 3.3 Supported Models; Table 2 model identity and Table 3 named row Version: 2506.14796v1 Retrieved: 2026-09-16T21:06:29.716004+00:00	source checked automated source review · 2026-09-23 Audit details Source review establishes this relationship only. Exact evaluated configurations and original numerical review status remain unchanged. Source identifies ESM-2 650M, ESM-C 600M and ESM3 as the respective foundation models evaluated with adapter tuning. Retain per-task configurations and metrics. Field: `links:family:discovery-model-esmc` Claim: model-evaluation-identity-7f6a71005656953611c1 Source artifact SHA-256: `59c7bbb888e8e91f33c1e2cabfde062c32381d6ca727d23b6655c977aabf97a2` Hash scope: Exact retrieved primary paper artifact bytes. Inspected artifact

All evidence (CSV)All evidence (JSONL)How to read the evidence

Sources and history

View linked audit checks and correction history

Release 2026-09-29-06401fd5b220 · Record review: source checked

1 source records and release history

pfmbench primary benchmark evidence · Original source · 2506.14796v1

Download this release

Technical metadata and extraction receipts

Stable ID: pfmbench-method-esm-c

areas: proteins-complexes
source locator: Table 3, row(ESM-C)
missing metadata: checkpoint revision: unreported; parameters: unextracted

Related records

family: ESMC
subject: ESM-C: family ESMC
model: ESM-C on PFMBench ANTI-RES: Antibiotic resistance
model: ESM-C on PFMBench BINDING-DB: BindingDB
model: ESM-C on PFMBench CLONING-CLF: Cloning CLF
model: ESM-C on PFMBench DEEPLOC2: DeepLoc2 Multi
model: ESM-C on PFMBench DEEPSOL: DeepSol
model: ESM-C on PFMBench EC: Enzyme Commission
model: ESM-C on PFMBench MAT-PROD: Material production
model: ESM-C on PFMBench METAL-ION: Metal ion binding
model: ESM-C on PFMBench PDB-BIND: PDBbind
model: ESM-C on PFMBench PROTEINGYM-ZS: ProteinGym zero-shot variant effect prediction
model: ESM-C on PFMBench SEC-STRUCT: Secondary structure
model: ESM-C on PFMBench STABILITY: TAPE_Stability

Suggest a correction