SAVRN
Search Contact SAVRN

Open-weight model · Fill mask

ESMC-6B

by Biohub biohub/ESMC-6B

ESMC is a state-of-the-art protein language model that has learned the rules of protein biology from training on billions of protein sequences.

Parameters6.4B
Context2,048
Weights25.4 GB
Licensemit
AccessOpen weights
Monthly Downloads2.9M

Runs On

What it takes to serve ESMC-6B (6.4B parameters): the memory its weights need at each precision, and the cheapest way to rent enough data-center GPUs to hold them.

PrecisionWeightsMemory neededCheapest setupPer hourAlso fits
16-bit 12.7 GB 15.2 GB 1x MI300X (192 GB)
Vultr		 $1.85 1x H100 $1.99 · 1x MI325X $2.00
8-bit 6.4 GB 7.6 GB 1x MI300X (192 GB)
Vultr		 $1.85 1x H100 $1.99 · 1x MI325X $2.00
4-bit 3.2 GB 3.8 GB 1x MI300X (192 GB)
Vultr		 $1.85 1x H100 $1.99 · 1x MI325X $2.00

Memory is the weights at that precision plus 20% for the runtime and a short context; a long context needs more. Prices are the lowest on-demand hourly rates in the SAVRN Index, read Sep 18, 2026.

Model Card

By Biohub, published under mit, revision af1602ba7406.

ESMC

Model Details

ESMC is a state-of-the-art protein language model that has learned the rules of protein biology from training on billions of protein sequences. ESMC provides representations of proteins enabling novel AI applications from therapeutic protein engineering to unlocking basic insights into protein biology across life.

The ESMC 6B model has 6 billion parameters, with 80 layers and 2.37e23 training flops. We additionally release overtrained 300M and 600M parameter variants of ESMC for local inference and finetuning.

The ESMFold2 structure prediction models are trained on top of a frozen ESMC 6B language model. ESMFold2 is a state-of-the-art model for protein structure prediction and design that defines a new frontier for speed and accuracy.

The ESMC sparse autoencoder, ESMC-6B-sae-layer60-k64-codebook16384, is built on the ESMC 6B model and provides human-interpretable, agent-generated feature descriptions. See the ESMC SAE overview card for the full set of ESMC SAE variants.

Read the full model card (1,376 words)

Configuration

Architecture
EsmcForMaskedLM
Context length (tokens)
2,048
Layers
80
Hidden size
2,560
Feed-forward size
6,912
Attention heads
40
Key/value heads
40
Head dimension
64
Vocabulary size
64
Model type
esmc

Identity and Version

Repository
biohub/ESMC-6B
Publisher
Biohub
Task
Fill mask
Modality
Text
Library
transformers
Parameters
6.4B parameters
Languages
en
Revision
af1602ba7406f521b11bf8f81d52af378cde09e4
First published
2026-05-19
Last updated
2026-09-16

Files and Weights

15 files, 25.4 GB in total. The weights are 6 files totalling 25.4 GB in safetensors.

Weights6 files · 25.4 GB
Configuration4 files · 92.1 KB
Tokenizer2 files · 3.2 KB
Documentation1 file · 15.0 KB
Other1 file · 30.2 KB
Repository1 file · 1.6 KB
Every file
FileTypeSizeSHA-256
model-00001-of-00006.safetensorsWeights4.9 GB 80cc4a51f218
model-00002-of-00006.safetensorsWeights5.0 GB 273740475dcd
model-00003-of-00006.safetensorsWeights5.0 GB 00769f0e680a
model-00004-of-00006.safetensorsWeights4.9 GB 1287dd34cb55
model-00005-of-00006.safetensorsWeights5.0 GB fef715f2b9f0
model-00006-of-00006.safetensorsWeights582.8 MB 7f88f2e9a884
config.jsonConfiguration1.1 KB —
model.safetensors.index.jsonConfiguration85.7 KB —
modeling_esmc_remote.pyConfiguration5.1 KB —
special_tokens_map.jsonConfiguration171 B —
README.mdDocumentation15.0 KB —
images/contact_pal.pngOther30.2 KB —
.gitattributesRepository1.6 KB —
tokenizer.jsonTokenizer2.9 KB —
tokenizer_config.jsonTokenizer327 B —

License and Download

License
mit
Access
Open weights, no gate
Download size
25.4 GB
Download from Biohub

Released by Biohub through its official repository on Hugging Face. Read the license.

Memory Requirements

PrecisionWeights in memory
As published25.4 GB
16-bit12.7 GB
8-bit6.4 GB
4-bit3.2 GB

Weights only, from the published parameter count; the key-value cache and runtime add to this.

Questions About ESMC-6B

How much GPU memory does ESMC-6B need?

About 15.2 GB at 16-bit and 3.8 GB at 4-bit: the weights (6.4B parameters) plus a working margin. A long context needs more.

What is the cheapest GPU to run ESMC-6B on?

At 16-bit, 1x MI300X from $1.85 an hour; at 4-bit, 1x MI300X from $1.85 an hour, at the lowest on-demand prices the SAVRN Index lists.

Can I use ESMC-6B commercially?

Yes. ESMC-6B is released under MIT License. The MIT License is a short permissive license. It permits commercial use, modification and redistribution, provided the copyright notice and permission notice are included.

What is ESMC-6B's context length?

2,048 tokens, from the maximum position embeddings in its published configuration.

Similar Models

Model · Fill mask

esm2_t33_650M_UR50D

AI at Meta

ESM-2 is a state-of-the-art protein model trained on a masked language modelling objective. It is suitable for fine-tuning on a wide range of tasks that take protein sequences as input. For detailed information on the model architecture and training data, please refer to the accompanying paper. You may also be interested in some demo notebooks (PyTorch, TensorFlow) which demonstrate how to fine-tune ESM-2 models on your tasks of interest. Several ESM-2 checkpoints are available in the Hub with varying sizes. Larger sizes generally have somewhat better accuracy, but require much more memory and time to train

Open weights mit 652M parameters 1,026 tokens transformers
View model

XLM-RoBERTa model pre-trained on 2.5TB of filtered CommonCrawl data containing 100 languages. It was introduced in the paper Unsupervised Cross-lingual Representation Learning at Scale by Conneau et al. and first released in this repository. Disclaimer: The team releasing XLM-RoBERTa did not write a model card for this model so this model card has been written by the Hugging Face team. XLM-RoBERTa is a multilingual version of RoBERTa. It is pre-trained on 2.5TB of filtered CommonCrawl data containing 100 languages. RoBERTa is a transformers model pretrained on a large corpus in a self-supervised fashion. This means it was pretrained on the raw texts only, with no humans labelling them in…

Open weights mit 561M parameters 514 tokens transformers
View model

Model · Fill mask

ModernBERT-large

Answer.AI

ModernBERT is a modernized bidirectional encoder-only Transformer model (BERT-style) pre-trained on 2 trillion tokens of English and code data with a native context length of up to 8,192 tokens. ModernBERT leverages recent architectural improvements such as: - Rotary Positional Embeddings (RoPE) for long-context support. - Local-Global Alternating Attention for efficiency on long inputs. - Unpadding and Flash Attention for efficient inference. ModernBERT’s native long context length makes it ideal for tasks that require processing long documents, such as retrieval, classification, and semantic search within large corpora. The model was trained on a large corpus of text and code, making it…

Open weights apache-2.0 396M parameters 8,192 tokens transformers
View model

Pretrained model on English language using a masked language modeling (MLM) objective. It was introduced in this paper and first released in this repository. This model is case-sensitive: it makes a difference between english and English. Disclaimer: The team releasing RoBERTa did not write a model card for this model so this model card has been written by the Hugging Face team. RoBERTa is a transformers model pretrained on a large corpus of English data in a self-supervised fashion. This means it was pretrained on the raw texts only, with no humans labelling them in any way (which is why it can use lots of publicly available data) with an automatic process to generate inputs and labels…

Open weights mit 355M parameters 514 tokens transformers
View model

Model · Fill mask

bert-large-uncased

BERT community

Pretrained model on English language using a masked language modeling (MLM) objective. It was introduced in this paper and first released in this repository. This model is uncased: it does not make a difference between english and English. Disclaimer: The team releasing BERT did not write a model card for this model so this model card has been written by the Hugging Face team. BERT is a transformers model pretrained on a large corpus of English data in a self-supervised fashion. This means it was pretrained on the raw texts only, with no humans labelling them in any way (which is why it can use lots of publicly available data) with an automatic process to generate inputs and labels from…

Open weights apache-2.0 336M parameters 512 tokens transformers
View model

This is a Japanese DeBERTa V2 large model pre-trained on Japanese Wikipedia, the Japanese portion of CC-100, and the Japanese portion of OSCAR. This model is trained with character-level tokenization and whole word masking. You can use this model for masked language modeling as follows: You can also fine-tune this model on downstream tasks. There is no need to tokenize texts in advance, and you can give raw texts to the tokenizer. The texts are tokenized into character-level tokens by sentencepiece. We used the following corpora for pre-training: - Japanese Wikipedia (as of 20221020, 3.2GB, 27M sentences, 1.3M documents) - Japanese portion of CC-100 (85GB, 619M sentences, 66M documents)…

Open weights cc-by-sa-4.0 330M parameters 512 tokens transformers
View model