LaMDA
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
- Organisation type
- Industry
- Country
- United States of America
- Published
- 10 February 2022
- Authors
- Romal Thoppilan, Daniel De Freitas, Jamie Hall, Noam Shazeer, Apoorv Kulshreshtha, Heng-Tze Cheng, Alicia Jin, Taylor Bos, Leslie Baker, Yu Du, YaGuang Li, Hongrae Lee, Huaixiu Steven Zheng, Amin Ghafouri, Marcelo Menegali, Yanping Huang, Maxim Krikun, Dmitry Lepikhin, James Qin, Dehao Chen, Yuanzhong Xu, Zhifeng Chen, Adam Roberts, Maarten Bosma, Yanqi Zhou, Chung-Ching Chang, Igor Krivokon, Will…
What it does
The problem areas the model was built for. A model can carry several of each.
- Domain
- Language
- Task
- Language modeling
- Approach
- Self-supervised learning
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
- 137B
- Training data
- 2,080,000,000,000 tokens
- Batch size
- 256,000
"LaMDA is a family of Transformer-based neural language models specialized for dialog, which have up to 137B parameters"
"and are pre-trained on 1.56T words of public dialog data and web text"
"All models were trained with 256K tokens per batch"
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
- 3.6 × 10²³ FLOP
- How it was established
- Hardware
"The total FLOPS is 56.5% * 123 TFLOPS/s * 1024 chips * 57.7 days = 3.55E+23" From https://arxiv.org/pdf/2201.08239.pdf p.18
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
- Google TPU v3
- Chips used
- 1,024
- Chip-hours
- 1,418,240
- Wall-clock time
- 1,385 hours (57.7 days)
- Hardware utilisation
- HFU 56.5%
- Power draw
- 927.2 kW
- Compute cost
- $229,950
57.7 days * 24
"We used the Lingvo framework [94] for training and achieved 123 TFLOPS/sec with 56.5% FLOPS utilization" HFU = 0.5650
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.
- Foundation model
- Yes
- Likely above 10²³ FLOP
- Yes
- Why it is tracked
- Historical significance
- Record confidence
- Confident
- Citations
- 1,863
Sources
Where this record came from and when it was last checked.
- Reference
- LaMDA: Language Models for Dialog Applications
- Last updated
- 25 May 2026
What the numbers mean
About this model
LaMDA was published by Google, in United States of America, in February 2022. It comes out of industry.
It works in Language, and is recorded as doing language modeling.
Because the weights are not available, none of the hardware figures elsewhere on this site apply to it.
How it was trained
Producing it required around 3.6 × 10²³ FLOP of arithmetic, on Google TPU v3, which is a statement about the training budget rather than about inference.
It was trained on about 2,080,000,000,000 tokens of text.
The reason it appears in this catalogue at all is historical significance.
Answers
LaMDA — common questions
How many parameters does LaMDA have?
LaMDA has 137B parameters. "LaMDA is a family of Transformer-based neural language models specialized for dialog, which have up to 137B parameters". 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.
Who created LaMDA?
LaMDA was published by Google, based in United States of America, categorised as industry.
When was LaMDA released?
LaMDA was published in February 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.
What is LaMDA used for?
LaMDA works in Language, and is recorded as handling language modeling. 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.
How much compute was used to train LaMDA?
Around 3.6 × 10²³ FLOP, on Google TPU v3. 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 LaMDA?
None. LaMDA 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 LaMDA open source?
No. LaMDA has not had its weights published, so it exists only as a service controlled by its owner.
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.