Check back regularly for additional model deployments.
Please reach out to request any new model deployments.
| Model | Slug | Tags | Tokenizer | Predictor | Classifier | de novo Generator |
Explainer | SemanticSimilarity | Creation Pipeline |
|---|---|---|---|---|---|---|---|---|---|
| AbLang-2 | ablang2 | antibody prediction generation embedding mlm | ✗ | - | |||||
| Design antibodies by predicting non-germline mutations that enhance specificity and affinity. Generate embeddings and predict mutation likelihoods with an antibody-specific language model optimized to overcome germline sequence bias. | |||||||||
| AbLang Heavy Chain API | ablang-heavy | antibody generator embeddings predict | ✗ | ✗ | - | ||||
| AbLang empowers researchers with capabilities like embeddings, sequence restoration, and likelihood computation, directly catering to the unique structural and functional attributes of heavy/light-chain antibodies. AbLang's deep learning framework, trained on an extensive collection of heavy-chain antibody sequences, offers unparalleled insights into their design and optimization for therapeutic applications. | |||||||||
| AbLang Light Chain API | ablang-light | antibody generator embeddings predict generate | ✗ | ✗ | - | ||||
| AbLang is an advanced AI language model tailored for antibody design, enabling embeddings, sequence restoration, and computing likelihoods for heavy/light-chain antibodies. Trained on a comprehensive database of human antibody sequences, AbLang can provide insights into antibody sequence functionality, and aid in the development of therapeutic antibodies. | |||||||||
| ABodyBuilder2 | abodybuilder2 | antibody structure prediction folding | ✗ | - | |||||
| Predict accurate antibody structures rapidly from sequence data. Generate structural models of antibody variable domains and CDR loops, with residue-level confidence scores, enabling large-scale immune protein analysis and characterization. | |||||||||
| ABodyBuilder3 Language | abodybuilder3-language | antibody structure prediction folding | ✗ | - | |||||
| Predict accurate 3D antibody structures from heavy/light chain sequences. Rapidly generate antibody-specific models optimized by deep learning, enabling high-throughput structural analysis of immune receptor datasets. | |||||||||
| ABodyBuilder3 pLDDT | abodybuilder3-plddt | antibody structure prediction folding | ✗ | - | |||||
| Predict accurate 3D antibody structures from heavy and light chain sequences, and obtain residue-level pLDDT confidence scores to assess model reliability and guide experimental validation strategies. | |||||||||
| AlphaFold2 | alphafold2 | protein structure folding alphafold2 msa jackhmmer mmseqs2 | ✗ | - | |||||
| Predict accurate 3D protein structures from amino acid sequences at near-experimental resolution. Leverage deep learning to analyze protein folding, model unknown structures, and accelerate structural biology research. | |||||||||
| AntiFold | antifold | protein antibody structure esm generation inverse-folding embedding inverse-fold | ✗ | ✗ | - | ||||
| Design antibody sequences from structure using inverse folding. Generate optimized CDR sequences predicted to maintain backbone structure, and predict mutation tolerance to guide affinity maturation and antibody optimization strategies. | |||||||||
| BioLMTox-2 | biolmtox2 | prediction embedding biosecurity toxins classification | ✗ | - | |||||
| Predict protein toxin potential rapidly and accurately using sequence-based embeddings. Screen peptides and proteins across multiple domains of life without alignment, supporting biosecurity, therapeutic development, and protein engineering. | |||||||||
| Chai-1 | chai1 | protein structure dna prediction esm transformer ligand multimer diffusion single-sequence multimodal rna | ✗ | - | |||||
| Predict accurate 3D structures for proteins, antibodies, multimers, nucleic acids, and protein-ligand complexes. Optionally integrate experimental constraints to enhance predictions for drug discovery applications. | |||||||||
| DNABERT-2 | dnabert2 | dna bert prediction embedding | ✗ | - | |||||
| Generate embeddings from DNA sequences or predict sequence log probabilities using DNABERT-2, a transformer model leveraging efficient Byte Pair Encoding (BPE) for accurate multi-species genome analysis with reduced computational complexity. | |||||||||
| DNABERT API | dnabert | dna bert | - | ||||||
| This pre-trained language model surpasses conventional methods, providing superior accuracy and insights for DNA sequence classification, variant calling, and more. Ideal for researchers aiming to unlock genomic complexities efficiently. | |||||||||
| DNA Chisel | dna-chisel | dna prediction analytics design feature-extraction | ✗ | - | |||||
| Model DNA sequences by generating dozens of sequence features and properties, like GC content, codon adaptation index, melting temperature, and restriction site counts. Increase the predictive potential of your models by incorporating these features into your workflows. | |||||||||
| ESM-1v | esm1v-all | prediction esm mlm | ✗ | - | |||||
| Predict the functional impact of protein sequence mutations with ESM-1v, a zero-shot transformer model. Score single or multiple amino acid substitutions to rapidly analyze mutational effects without additional experimental data or training. | |||||||||
| ESM-1v #1 API/v1 | esm1v_t33_650M_UR90S_1 esm1v_t33_650M_UR90S_2 esm1v_t33_650M_UR90S_3 esm1v_t33_650M_UR90S_4 esm1v_t33_650M_UR90S_5 |
enzyme protein maturation esm | ✗ | - | |||||
| First of five models in the ESM-1v series, each trained with a different seed. Predict favorable or unfavorable point-variants. NLP model trained on UniRef90. Zero-shot (unsupervised) predictor of functional effects. Ensemble with the remaining four models for best results. doi | |||||||||
| ESM1v ALL API | esm1v-all | enzyme protein maturation esm | ✗ | - | |||||
| Conveniently retrieve predictions from all five ESM-1v models for a given position, allowing for rapid variant ranking and deep mutational scans. | |||||||||
| ESM1v N1 API | esm1v-n1 | enzyme protein maturation esm | ✗ | - | |||||
| The 1st of the ESM-1v models, for unmasking AA positional likelihoods. | |||||||||
| ESM1v N2 API | esm1v-n2 | enzyme protein maturation esm | ✗ | - | |||||
| 2nd of the ESM-1v models. | |||||||||
| ESM1v N3 API | esm1v-n3 | enzyme protein maturation esm | ✗ | - | |||||
| 3rd ESM-1v model. | |||||||||
| ESM1v N4 API | esm1v-n4 | enzyme protein maturation esm | ✗ | - | |||||
| 4th ESM-1v model. | |||||||||
| ESM1v N5 API | esm1v-n5 | enzyme protein maturation esm | ✗ | - | |||||
| The 5th of the ESM-1v models, for unmasking AA positional likelihoods. | |||||||||
| ESM-2 150M | esm2-150m | protein prediction esm embedding mlm | ✗ | - | |||||
| Generate embeddings capturing evolutionary and structural insights directly from protein sequences. Predict residue-residue contacts and infer protein structure at atomic resolution without external evolutionary databases or alignments. | |||||||||
| ESM-2 35M | esm2-35m | protein prediction esm embedding mlm | ✗ | - | |||||
| Generate embeddings for protein sequences or predict masked amino acids using the compact ESM-2 35M language model. Efficiently analyze protein properties and evolutionary patterns directly from sequence data. | |||||||||
| ESM2 3B API | esm2-3b | enzyme protein embeddings esm logits | ✗ | - | |||||
| Use this 2nd-largest ESM-2 model for highly informative embeddings, contact maps, and more. | |||||||||
| ESM-2 650M | esm2-650m | enzyme protein antibody prediction esm transformer embedding mlm | ✗ | - | |||||
| Generate embeddings from protein sequences or predict masked amino acids using ESM-2 650M, a protein language model capturing evolutionary patterns to inform structure and function, suitable for protein analysis and design tasks. | |||||||||
| ESM-2 8M | esm2-8m | protein prediction esm embedding mlm | ✗ | - | |||||
| Generate embeddings and predict masked amino acids in protein sequences using a compact transformer-based language model trained on evolutionary patterns, enabling rapid analysis of protein structure and function. | |||||||||
| ESM C 300M | esmc-300m | protein prediction esm language-model embedding mlm | ✗ | - | |||||
| Generate embeddings from protein sequences using ESM C 300M, a transformer-based model trained on evolutionary-scale data. Predict protein properties and inform downstream design and analysis tasks with high efficiency and accuracy. | |||||||||
| ESM C 600M | esmc-600m | protein prediction esm embedding mlm | ✗ | - | |||||
| Generate biologically meaningful embeddings for protein sequences and predict masked residues with ESM C 600M, a transformer-based model trained on evolutionary-scale data for accurate representation learning and functional protein analysis. | |||||||||
| ESMFold | esmfold | protein structure prediction esm folding alphafold2 | ✗ | - | |||||
| Predict accurate 3D atomic-level structures of single or multiple protein chains directly from sequence using ESMFold. Obtain confidence scores to assess model reliability and support protein design and functional analysis workflows. | |||||||||
| ESMFold Multi-Chain API | esmfold-multichain | protein structure prediction esm | ✗ | - | |||||
| For multi-chain proteins like antibodies, predict the folded structure in seconds with v2 of ESMFold. Similar results, speed, and accuracy to single-chain folding endpoint, but now available for more complex sequences. PDBs via REST in less than a minute, using one of the largest protein LMs to date. doi | |||||||||
| ESM-IF1 | esm-if1 | structure esm generation inverse-fold | ✗ | - | |||||
| Generate protein sequences from backbone coordinates using the ESM-IF1 inverse folding model. Design novel proteins, protein complexes, and binding interfaces by predicting sequences compatible with desired 3D structural conformations. | |||||||||
| Evo 1.5 8k Base | evo-15-8k-base | protein dna prediction generation multimodal rna | ✗ | ✗ | - | ||||
| Generate and predict DNA sequences at genome scale with Evo 1.5 8k Base, an autoregressive genomic foundation model trained on prokaryotic genomes. Design multi-gene systems, CRISPR-Cas complexes, and analyze nucleotide-level gene essentiality. | |||||||||
| Evo 2 1B Base | evo2-1b-base | protein dna prediction generation multimodal rna embedding | ✗ | ✗ | - | ||||
| Generate and analyze DNA sequences at genome scale with Evo 2 1B Base, a genomic language model trained across all domains of life. Predict mutational effects and design novel DNA sequences with accurate evolutionary context and biological coherence. | |||||||||
| Evo v1.5-8k Base | evo-v1.5-8k | dna rna language-model | ✗ | ✗ | - | ||||
| 1.5-B parameter Evo multimodal LM (DNA/RNA) with 8 k context; supports log-prob scoring and sequence generation. | |||||||||
| Global Label Membrane MPNN | global-label-membrane-mpnn | protein design mpnn | ✗ | - | |||||
| Global Label Membrane MPNN – membrane-aware variant that tailors amino-acid choices based on a coarse 'membrane' vs 'soluble' flag for the whole chain, streamlining transmembrane protein design. | |||||||||
| IgBert Paired | igbert-paired | antibody embeddings bert prediction generation language-model embedding | ✗ | ✗ | - | ||||
| Generate embeddings from paired antibody sequences, predict sequence probabilities, or design novel antibody candidates leveraging IgBert's antibody-specific language model trained on extensive paired heavy-light chain sequence data. | |||||||||
| IgBert Unpaired | igbert-unpaired | antibody bert prediction generation embedding | ✗ | ✗ | - | ||||
| Generate embeddings of antibody variable region sequences with IgBert Unpaired, an antibody-specific transformer language model trained on over two billion sequences, enabling accurate analysis and prediction for antibody engineering tasks. | |||||||||
| IgT5 Paired | igt5-paired | antibody embeddings t5 embedding | - | ||||||
| Generate embeddings of paired antibody variable region sequences using the IgT5 Paired algorithm. Predict binding affinity and support antibody engineering tasks leveraging cross-chain features learned from extensive paired antibody data. | |||||||||
| IgT5 Unpaired | igt5-unpaired | antibody embeddings t5 embedding | - | ||||||
| Generate embeddings from unpaired antibody variable region sequences using IgT5, an antibody-specific language model trained on over two billion sequences, to analyze antibody properties and inform therapeutic antibody design. | |||||||||
| ImmuneFold Antibody | immunefold-antibody | antibody structure prediction folding | ✗ | - | |||||
| Predict accurate 3D antibody structures from sequences, leveraging parameter-efficient transfer learning tailored for immune proteins. Generate precise structural predictions of hypervariable regions to guide antibody design and immunotherapy research. | |||||||||
| ImmuneFold TCR | immunefold-tcr | structure prediction folding tcr t-cell | ✗ | - | |||||
| Predict accurate 3D structures of T-cell receptors (TCRs) using a LoRA-fine-tuned Evoformer architecture. Leverage predictions for zero-shot assessment of TCR-epitope binding, aiding immunotherapy research and TCR engineering applications. | |||||||||
| Ligand ProteinMPNN | ligand-mpnn | protein design mpnn | ✗ | - | |||||
| Ligand ProteinMPNN – incorporates bound ligand atoms and optional side-chain repacking to optimise sequences for small-molecule or cofactor binding sites. | |||||||||
| MPNN | mpnn | protein design mpnn | ✗ | - | |||||
| ProteinMPNN generative design model. Supports sequence generation from input PDBs. | |||||||||
| nanoBERT | nanobert | antibody embeddings bert prediction generation language-model embedding nanobody | ✗ | ✗ | - | ||||
| Design nanobody variants by predicting biologically feasible amino acid substitutions. Use nanoBERT's transformer-based embeddings to explore mutational space and optimize therapeutic nanobody stability and function. | |||||||||
| NanoBodyBuilder2 | nanobodybuilder2 | antibody structure prediction folding nanobody | ✗ | - | |||||
| Predict accurate 3D structures of nanobody heavy chain sequences rapidly. Generate precise CDR loop conformations and obtain residue-level error estimates, enabling efficient structural analysis of nanobody repertoires. | |||||||||
| Omni-DNA 1B | omni-dna-1b | dna embeddings prediction multimodal language-model embedding embedding print(resp.json()) | ✗ | - | |||||
| Encode DNA sequences into embeddings or predict sequence log probabilities using Omni-DNA 1B, a unified, transformer-based genomic model for cross-modal and multi-task learning across diverse genomic annotation and prediction tasks. | |||||||||
| Peptides | peptides | protein antibody prediction feature-extraction peptide | - | ||||||
| Generate features from peptide and protein sequences related to charge, hydrophobicity, stability, molecular weight, and other biophysical/physicochemical properties. Inform the design of novel proteins and peptides based on computed physicochemical and structural properties. | |||||||||
| Per-Residue Label Membrane MPNN | per-residue-label-membrane-mpnn | protein design mpnn | ✗ | - | |||||
| Per-Residue Label Membrane MPNN – allows per-residue transmembrane and interface annotations, enabling precise control over buried, interface, and solvent-exposed segments of membrane proteins. | |||||||||
| ProGen2 BFD90 | progen2-bfd90 | ✗ | - | ||||||
| ProGen2 variant trained on BFD90 dataset, excelling in diverse protein design applications. | |||||||||
| ProGen2 BFD90 API | progen2-bfd90 | protein generator gpt | ✗ | - | |||||
| BFD-90 pretrained model from the suite of ProGen2 generative models. Tune outputs with your choice of pretrained model, temperature, length, and more. | |||||||||
| ProGen2 Large | progen2-large | ✗ | - | ||||||
| A 2.7B parameter model providing enhanced capabilities for complex protein generation tasks. | |||||||||
| ProGen2 Large API | progen2-large | protein generator gpt | ✗ | - | |||||
| ProGen2 Large contains 2.7B parameters and is the second-largest model in the suite of generative models. Tune outputs with your choice of pretrained model, temperature, length, and more. | |||||||||
| ProGen2 Medium | progen2-medium | ✗ | - | ||||||
| A 764M parameter model offering a balance between computational efficiency and performance for general protein generation tasks. | |||||||||
| ProGen2 Medium API | progen2-medium | protein generator gp2 | ✗ | - | |||||
| For faster protein generation, use this 764M parameter model from the ProGen2 suite of generative models. Tune outputs with your choice of pretrained model, temperature, length, and more. | |||||||||
| ProGen2 OAS | progen2-oas | ✗ | - | ||||||
| ProGen2 variant trained on OAS sequences, specializing in antibody generation. | |||||||||
| ProGen2 OAS API | progen2-oas | antibody generator gpt | ✗ | - | |||||
| ProGen2 OAS was trained on the Observed Antibody Space, making it more suitable to some generative-antibody applications. Tune outputs with your choice of pretrained model, temperature, length, and more. | |||||||||
| ProstT5 AA2Fold | prostt5-aa2fold | enzyme protein antibody structure embeddings folding proteins generation seq2seq encoder-decoder inverse-folding t5 embedding | ✗ | - | |||||
| Predict accurate 3Di structure sequences directly from amino acid inputs. ProstT5 AA2Fold generates embeddings and fold predictions enabling rapid protein structure-based homology searches without costly 3D modeling. | |||||||||
| ProstT5 Fold2AA | prostt5-fold2aa | enzyme protein antibody structure embeddings folding proteins generation seq2seq encoder-decoder inverse-folding t5 embedding inverse-fold | ✗ | - | |||||
| Generate amino acid sequences from protein 3Di structural tokens. Leverage structure embeddings to rapidly identify remote protein homologs and analyze structural relationships without explicit 3D prediction. | |||||||||
| ProteInfer EC (Enzyme Commmision) API | proteinfer-ec | enzyme protein prediction EC | ✗ | - | |||||
| ProteInfer EC Prediction API leverages AI to accurately predict enzyme commission (EC) numbers from protein sequences, enhancing enzymatic research and biotechnological applications. | |||||||||
| ProteInfer GO (Gene Ontology) API | proteinfer-go | protein GO predictor function | ✗ | - | |||||
| Harness the power of deep learning to predict Gene Ontology (GO) terms for proteins directly from sequence data. ProteInfer leverages a comprehensive dataset and advanced models to provide accurate functional annotations, offering insights into biological processes, cellular components, and molecular functions. | |||||||||
| ProteinMPNN | protein-mpnn | protein design mpnn | ✗ | - | |||||
| ProteinMPNN – a message-passing neural-network for protein sequence design. Trained on >25 k PDB structures, it achieves ~52 % native sequence recovery and has been validated experimentally (Science, 2022). Generates sequences in ≈1 s for a 100-residue backbone. | |||||||||
| Sadie Antibody | sadie-antibody | antibody prediction classification feature-extraction nanobody renumbering | ✗ | - | |||||
| Annotate antibody sequences using the SADIE algorithm to identify framework and CDR regions, germline assignments, and sequence numbering. Generate AIRR-compliant data outputs for immunoinformatics and antibody discovery workflows. | |||||||||
| Soluble ProteinMPNN | soluble-mpnn | protein design mpnn | ✗ | - | |||||
| Soluble ProteinMPNN – fine-tuned to favour hydrophilic surface residues, improving soluble expression yields while retaining the speed and accuracy of the baseline model. | |||||||||
| TCRBuilder2 | tcrbuilder2 | structure prediction folding tcr t-cell | ✗ | - | |||||
| Predict accurate 3D T-cell receptor (TCR) structures from alpha and beta chain sequences. Rapidly generate reliable models to analyze TCR-antigen interactions and facilitate therapeutic TCR design and engineering efforts. | |||||||||
| TCRBuilder2+ | tcrbuilder2-plus | structure prediction folding tcr t-cell | ✗ | - | |||||
| Predict accurate 3D structures of T-cell receptors from paired alpha/beta chain sequences. Rapidly analyze receptor-antigen interactions and generate reliable structural models for immunotherapy development and immune repertoire studies. | |||||||||
| UniRef50 Embedding Similars API | uniref50-similars | ✗ | - | ||||||
| Accelerate your protein sequence searches with our blazing-fast Nearest Neighbors Search API. By leveraging protein language models to generate vector embeddings for over 65 million UniRef50 sequences, our service enables lightning-fast similarity searches that are 1,200 times faster than traditional Levenshtein-based methods like BLAST. | |||||||||
| ZymCTRL API | zymctrl | enzymes generation | ✗ | - | |||||
Create Your Own With
| Language Model | Slug | Classifier | Regressor | Generator | Tags |
|---|---|---|---|---|---|
| Protein Classifier | finetune_esm2_classifier | ✗ | protein classifier | ||
| Start with one of the largest protein language models, trained on hundreds of millions of proteins. Then finetune it on your own sequences, to make it specific to your classification task. Creates a GPU-backed API endpoint for classifying new sequences, retrieving probabilities, with your personalized ESM2 model. doi | |||||
| Protein Generator | finetune_protgpt2_generator | ✗ | protein generator | ||
| Finetune a generative protein model on your own data, then create new sequences via API. Create de-novo best representative sequences that "follow the principles of natural ones and sample unexplored protein space". Only positive-class sequences required. doi | |||||
| DNA Classifier | finetune_dnabert_classifier | ✗ | dna bert | ||
| DNA-based BERT models have demonstrated performance on promoter prediction, splice sites and transcription factor binding sites, up to 99% AUC. Bring your own NGS data or relevant public datasets to create any classifier. Your model also comes with an ability to export meaningful DNA BERT embeddings, perform semantic similarity searches. doi | |||||