OmegaPLM

Closed weights Massachusetts Institute of Technology (MIT),Westlake University 670M parameters July 2022

No estimate

No hardware requirements for this model

The weights for this model have not been published, so it cannot be downloaded or run on your own hardware at any size. It is reachable only through its provider, and no graphics card changes that.

On record

Full specification

Everything on record for this model. Most of it describes how it was trained rather than how it runs — useful context for judging how much work went into it, and how it compares with models built at a different scale.

Origin

Who built this model, where, and when it was published.

Organisation
Massachusetts Institute of Technology (MIT),Westlake University
Organisation type
Academia,Academia
Country
United States of America, China
Published
22 July 2022
Authors
Ruidong Wu, Fan Ding, Rui Wang, Rui Shen, Xiwen Zhang, Shitong Luo, Chenpeng Su, Zuofan Wu, Qi Xie, Bonnie Berger, Jianzhu Ma, Jian Peng

What it does

The problem areas the model was built for. A model can carry several of each.

Domain
Biology
Task
Proteins, Protein folding prediction
Numerical format
TF32

Size

How large the model is and how much data it was trained on. Parameters are the figure that decides whether it fits on a given graphics card.

Parameters
670M

"Our model contains 66 layers with around 670 million parameters without sharing parameters, which doubles the layer count of ESM-1b but roughly retains the parameter count."

Training data
1,258,291,200,000 tokens

Number of sequences: 35 x 10^6 Sequence length: 512 Total data points: 35 x 10^6 x 512 = 1.792 x 10^10 tokens First stage crop size: 256 First stage data points: 35 x 10^6 x 256 = 8.96 x 10^9 tokens Additional structural data: ~110,000 sequences = 7.68 x 10^7 tokens Final estimate: 8.96 x 10^9 tokens

Batch size
2,097,152

"[...] each batch contains 4,096 sequences and each sequence is padded or cropped to 512 residues" 4096 * 512 = 2097152

Training compute

The arithmetic performed to train the model, measured in floating-point operations. It is a measure of what the training run cost, not of how fast the finished model answers you.

Training compute
1 × 10²² FLOP

"OmegaPLM is implemented in PyTorch (44) and trained for 2,560 GPU Nvidia A100 80G days." "Default precision format in Nvidia A100 GPUs is set to TensorFloat-32 for matrix operations." Assume 0.3 utilization for language model Estimate: (2560 * 24 * 3600) s * 156e12 FLOP/s * 0.3 * = 1.04e22

How it was established
Hardware

The training run

What it physically took to train: which chips, how many, for how long, and what that drew from the wall.

Training hardware
NVIDIA A100 SXM4 80 GB
Chip-hours
61,440
Compute cost
$52,400

Availability

Whether you can obtain the model and run it on your own hardware, which is what decides if any of the graphics-card figures on this page apply.

Weights
Closed — provider access only
Model access
Unreleased
Training code
Unreleased

How it is classified

Labels the source dataset applies when tracking notable models, and how confident it is in the entry.

Why it is tracked
Historical significance

"Here, we introduce OmegaFold, the first computational method to successfully predict high-resolution protein structure from a single primary sequence alone. Using a new combination of a protein language model that allows us to make predictions from single sequences and a geometry-inspired transformer model trained on protein structures, OmegaFold outperforms RoseTTAFold and achieves similar prediction accuracy to AlphaFold2 on recently released structures"

Record confidence
Confident
Citations
445

Sources

Where this record came from and when it was last checked.

Reference
High-resolution de novo structure prediction from primary sequence
Last updated
1 January 2026

What the numbers mean

About this model

OmegaPLM was published by Massachusetts Institute of Technology (MIT),Westlake University, in United States of America, in July 2022. The organisation is categorised as academia,Academia.

It works in Biology, and is recorded as doing proteins, Protein folding prediction.

This is a closed model: the trained values stayed with whoever produced them, and there is no local version to run.

What went into building it

Producing it required around 1 × 10²² FLOP of arithmetic, on NVIDIA A100 SXM4 80 GB, which is a statement about the training budget rather than about inference.

It was trained on about 1,258,291,200,000 tokens of text.

The reason it appears in this catalogue at all is historical significance.

Answers

OmegaPLM — common questions

01

How much compute was used to train OmegaPLM?

Around 1 × 10²² FLOP, on NVIDIA A100 SXM4 80 GB. That measures what producing the model cost and says nothing about how quickly it answers once trained — inference speed comes from memory bandwidth, not from the training budget.

02

What GPU do I need to run OmegaPLM?

None. OmegaPLM is a closed model — its weights were never published, so it cannot be downloaded or run on your own hardware at any price. It is reachable only through its provider.

03

Is OmegaPLM open source?

No. OmegaPLM has not had its weights published, so it exists only as a service controlled by its owner.

04

How many parameters does OmegaPLM have?

OmegaPLM has 670M parameters. "Our model contains 66 layers with around 670 million parameters without sharing parameters, which doubles the layer count of ESM-1b but roughly retains the parameter count.". That figure is the total, and it is what decides how much memory the model needs — roughly half a gigabyte per billion at the compression most people use.

05

Who created OmegaPLM?

OmegaPLM was published by Massachusetts Institute of Technology (MIT),Westlake University, based in United States of America, categorised as academia,Academia.

06

When was OmegaPLM released?

OmegaPLM was published in July 2022. Capability per parameter has improved considerably since, so a newer model of the same size is often the better use of the same hardware.

07

What is OmegaPLM used for?

OmegaPLM works in Biology, and is recorded as handling proteins, Protein folding prediction. These are the areas it was designed around; they describe intent rather than a hard boundary.

Source

Original publication

Record last updated 1 January 2026

The other direction

Looking at it from the other side?

This page starts from the model. If you already own a card and want to know everything it will run, start from the hardware instead.