ESM-AA

Closed weights Peking University,Nanjing University,Tsinghua University,PharMolix 35M parameters April 2024

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
Peking University,Nanjing University,Tsinghua University,PharMolix
Organisation type
Academia,Academia,Academia,Industry
Country
China
Published
5 April 2024
Authors
Kangjie Zheng, Siyu Long, Tianyu Lu, Junwei Yang, Xinyu Dai, Ming Zhang, Zaiqing Nie, Wei-Ying Ma, Hao Zhou

What it does

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

Domain
Biology
Task
Protein folding prediction, Protein or nucleotide language model (pLM/nLM), Proteins
Base model
ESM2-35M

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
35M
Training data
1,143,750,000 tokens

Assuming 300 tokens per protein and 100 token per molecule 8M*300+209M*100=23300000000

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
7.3 × 10²⁰ FLOP

1. Hardware: 16x NVIDIA A100 GPUs (3.12e14 FLOP/s per GPU) 2. Training duration: 3 days directly reported = 259,200 seconds 3. Utilization: 40% assumed 4. Calculation: 3.12e14 FLOP/s × 16 GPUs × 259,200s × 0.40 = 5.18e20 FLOPs Base model: 209999999999999970000 Total: 209999999999999970000+5.18e20=728000000000000000000

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
Chips used
16
Wall-clock time
72 hours
Power draw
12.7 kW

How it is classified

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

Record confidence
Confident
Citations
19

Sources

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

Reference
ESM All-Atom: Multi-scale Protein Language Model for Unified Molecular Modeling
Last updated
25 May 2026

What the numbers mean

Where it came from

ESM-AA was published by Peking University,Nanjing University,Tsinghua University,PharMolix, in the country recorded as China, during April 2024. The category the publisher falls under is academia,Academia,Academia,Industry.

It works in the domain of Biology, and is recorded as performing the task of protein folding prediction, Protein or nucleotide language model (pLM/nLM), Proteins.

Its starting point was an existing base model, ESM2-35M. That is why it shares the base model's general shape and size.

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

How it was trained

The training run consumed about 7.3 × 10²⁰ FLOP, on hardware recorded as NVIDIA A100. That figure measures what producing the model cost, and has no bearing on how fast it answers.

It was trained on a corpus of about 1,143,750,000 tokens of text.

Answers

ESM-AA — common questions

01

ESM-AA— how many parameters does it have?

It has a parameter count of 35M. 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.

02

ESM-AA— who created it?

It was published by Peking University,Nanjing University,Tsinghua University,PharMolix, based in China, an organisation categorised as academia,Academia,Academia,Industry.

03

ESM-AA— when was it released?

It was published in April 2024. Capability per parameter has improved considerably since, so a newer model of the same size is often the better use of the same hardware.

04

ESM-AA— what is it used for?

It works in the domain of Biology, and is recorded as handling the task of protein folding prediction, Protein or nucleotide language model (pLM/nLM), Proteins. These are the areas it was designed around; they describe intent rather than a hard boundary.

05

ESM-AA— how much compute was used to train it?

Training consumed around 7.3 × 10²⁰ FLOP, on hardware recorded as NVIDIA A100. 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.

06

ESM-AA— what GPU do I need to run it?

None. This 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.

07

ESM-AA— is it open source?

The licensing was never recorded in our source data. We treat unstated licensing as closed, because an unrecorded licence is not one to rely on.

Source

Original publication

Record last updated 25 May 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.