ReALM
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
- Apple
- Organisation type
- Industry
- Country
- United States of America
- Published
- 29 March 2024
- Authors
- Joel Ruben Antony Moniz, Soundarya Krishnan, Melis Ozyildirim, Prathamesh Saraf, Halim Cagri Ates, Yuan Zhang, Hong Yu, Nidhi Rajshree
What it does
The problem areas the model was built for. A model can carry several of each.
- Domain
- Language
- Task
- Named entity recognition (NER), Language modeling, Part-of-speech tagging
- Approach
- Supervised
- Base model
- Flan-T5 11B
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
- 3B
- Training data
- 134,000,000,000 tokens
Fine-tuned FLAN-T5 models ranging from 80M to 3B
2300 training examples from conversation; 3900 synthetically generated training examples; 10100 training examples using context from a phone screen.
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
- SOTA improvement
- Record confidence
- Confident
"We show that ReaLM outperforms previous approaches, and performs roughly as well as the state-of-the-art LLM today, GPT-4, despite consisting of far fewer parameters." "We also benchmark against GPT-3.5 and GPT-4, with our smallest model achieving performance comparable to that of GPT-4, and our larger models substantially outperforming it." I don't see any standard benchmarks that they would claim SOTA on
Sources
Where this record came from and when it was last checked.
- Reference
- ReALM: Reference Resolution As Language Modeling
- Last updated
- 28 November 2025
What the numbers mean
Where it came from
ReALM was published by Apple, in United States of America, in March 2024. It comes out of industry.
It works in Language, and is recorded as doing named entity recognition (NER), Language modeling, Part-of-speech tagging.
It builds on Flan-T5 11B, which is why it shares that 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.
What went into building it
It was trained on about 134,000,000,000 tokens of text.
It is tracked in the underlying dataset for one reason in particular: sOTA improvement.
Answers
ReALM — common questions
Who created ReALM?
ReALM was published by Apple, based in United States of America, categorised as industry.
When was ReALM released?
ReALM was published in March 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 ReALM used for?
ReALM works in Language, and is recorded as handling named entity recognition (NER), Language modeling, Part-of-speech tagging. 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.
What GPU do I need to run ReALM?
None. ReALM 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 ReALM open source?
No. ReALM has not had its weights published, so it exists only as a service controlled by its owner.
How many parameters does ReALM have?
ReALM has 3B parameters. Fine-tuned FLAN-T5 models ranging from 80M to 3B. 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.
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.