Mothra
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
- Tokyo Institute of Technology
- Organisation type
- Academia
- Country
- Japan
- Published
- 25 September 2024
- Authors
- Takamasa Suzuki, Dian Ma, Nobuaki Yasuo, Masakazu Sekijima
What it does
The problem areas the model was built for. A model can carry several of each.
- Domain
- Biology
- Task
- 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
250,000 molecules × 50 tokens/molecule = 12,500,000 tokens Total training datapoints = 12,500,000 tokens (1.25e7)
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
- 3.7 × 10¹⁹ FLOP
- How it was established
- Hardware
1. Hardware setup: 4x NVIDIA Tesla P100 GPUs (1.90×10¹³ FLOP/s per GPU) 2. Training duration: 14 days (directly provided) = 1,209,600 seconds 3. Utilization rate: 40% 4. Final calculation: 1.90×10¹³ FLOP/s/GPU × 4 GPUs × 1,209,600 seconds × 0.4 = 3.7×10¹⁹ FLOPs
How it is classified
Labels the source dataset applies when tracking notable models, and how confident it is in the entry.
- Record confidence
- Likely
Sources
Where this record came from and when it was last checked.
- Reference
- Mothra: Multiobjective de novo Molecular Generation Using Monte Carlo Tree Search
- Last updated
- 28 November 2025
What the numbers mean
What this model is
Mothra was published by Tokyo Institute of Technology, in Japan, in September 2024. The organisation is categorised as academia.
It works in Biology, and is recorded as doing drug discovery.
Its weights were never published, so it can only be reached through its provider. No graphics card changes that.
How it was trained
The training run consumed about 3.7 × 10¹⁹ FLOP. That figure describes the cost of creating it and has no bearing on how quickly it generates text.
Answers
Mothra — common questions
How much compute was used to train Mothra?
Around 3.7 × 10¹⁹ FLOP. 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.
What GPU do I need to run Mothra?
None. Mothra 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.
Is Mothra open source?
The licensing for Mothra was never recorded in our source data. We treat unstated licensing as closed, because an unrecorded licence is not one to rely on.
How many parameters does Mothra have?
No parameter count has been published for Mothra, which is why no memory or speed figure appears on this page.
Who created Mothra?
Mothra was published by Tokyo Institute of Technology, based in Japan, categorised as academia.
When was Mothra released?
Mothra was published in September 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.
What is Mothra used for?
Mothra works in Biology, and is recorded as handling drug discovery. A model can carry several of each, so these are the areas it was built for rather than a limit on what it will attempt.
Source
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.