Uni-Mol Molecular Model

Open weights Renmin University of China,DP Technology,AI for Science Institute, Beijing (AISI) March 2023

No estimate

No hardware requirements for this model

This model's weights are open, but no parameter count has been published for it. Every memory and speed figure starts from that number, so we would rather show nothing than a fabricated estimate.

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
Renmin University of China,DP Technology,AI for Science Institute, Beijing (AISI)
Organisation type
Academia,Industry,Government
Country
China
Published
6 March 2023
Authors
Gengmo Zhou, Zhifeng Gao, Qiankun Ding, Hang Zheng, Hongteng Xu, Zhewei Wei, Linfeng Zhang, Guolin Ke

What it does

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

Domain
Biology
Task
Molecular representation learning, Molecular property prediction, Protein-ligand contact prediction, Drug discovery

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.

Training data
tokens

"a molecular model pretrained by 209M molecular conformations" Table 6: batch size 128 Max training steps 1M

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
5.4 × 10¹⁸ FLOP

31330000000000 FLOP / sec / GPU [fp16 asssumed] * 8 GPUs * 20 hours * 3600 sec / hour * 0.3 [assumed utilization] = 5.413824e+18 FLOP

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 Tesla V100 DGXS 32 GB
Chips used
8
Wall-clock time
20 hours

"Molecular pretraining runs on 8 V100 GPUs (32GB memory, the same below), and the training time is about 20 hours."

Power draw
4.0 kW

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
Open — downloadable
Model access
Open weights (unrestricted)
Training code
Open source

https://github.com/deepmodeling/Uni-Mol MIT license

How it is classified

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

Record confidence
Confident

Sources

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

Reference
Uni-Mol: A Universal 3D Molecular Representation Learning Framework
Last updated
28 November 2025

What the numbers mean

What this model is

Uni-Mol Molecular Model was published by Renmin University of China,DP Technology,AI for Science Institute, Beijing (AISI), in China, in March 2023. The organisation is categorised as academia,Industry,Government.

It works in Biology, and is recorded as doing molecular representation learning, Molecular property prediction, Protein-ligand contact prediction, Drug discovery.

Because its weights were released, nothing about running it depends on a provider staying available — it is yours once downloaded.

Training and provenance

The training run consumed about 5.4 × 10¹⁸ FLOP, on NVIDIA Tesla V100 DGXS 32 GB. That figure describes the cost of creating it and has no bearing on how quickly it generates text.

Answers

Uni-Mol Molecular Model — common questions

01

What is Uni-Mol Molecular Model used for?

Uni-Mol Molecular Model works in Biology, and is recorded as handling molecular representation learning, Molecular property prediction, Protein-ligand contact prediction, Drug discovery. Models frequently carry more than one of each, and the tags describe purpose rather than capability limits.

02

Where can I download Uni-Mol Molecular Model?

The weights for Uni-Mol Molecular Model are published, though we do not hold a repository link for it. This site calculates hardware requirements rather than hosting model files.

03

How much compute was used to train Uni-Mol Molecular Model?

Around 5.4 × 10¹⁸ FLOP, on NVIDIA Tesla V100 DGXS 32 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.

04

What GPU do I need to run Uni-Mol Molecular Model?

We cannot say. Uni-Mol Molecular Model has open weights, but no parameter count has been published for it, and every memory and speed calculation starts from that number. We would rather show nothing than a fabricated estimate.

05

Is Uni-Mol Molecular Model open source?

Its weights are published, so Uni-Mol Molecular Model can be downloaded and run on your own hardware. Note that open weights is not the same as open source in the full sense — it says nothing about the training data, the training code, or the commercial terms attached.

06

How many parameters does Uni-Mol Molecular Model have?

No parameter count has been published for Uni-Mol Molecular Model, which is why no memory or speed figure appears on this page.

07

Who created Uni-Mol Molecular Model?

Uni-Mol Molecular Model was published by Renmin University of China,DP Technology,AI for Science Institute, Beijing (AISI), based in China, categorised as academia,Industry,Government.

08

When was Uni-Mol Molecular Model released?

Uni-Mol Molecular Model was published in March 2023. Capability per parameter has improved considerably since, so a newer model of the same size is often the better use of the same hardware.

Source

Original publication

Record last updated 28 November 2025

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.