GPT-4.5

Closed weights OpenAI February 2025

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
OpenAI
Organisation type
Industry
Country
United States of America
Published
27 February 2025
Authors
Foundational contributors Alex Paino, Ali Kamali, Amin Tootoonchian, Andrew Tulloch, Ben Sokolowsky, Clemens Winter, Colin Wei, Daniel Kappler, Daniel Levy, Felipe Petroski Such, Geoff Salmon, Ian O’Connell, Jason Teplitz, Kai Chen, Nik Tezak, Prafulla Dhariwal, Rapha Gontijo Lopes, Sam Schoenholz, Youlong Cheng, Yujia Jin, Yunxing Dai Research Core contributors Aiden Low, Alec Radford, Alex Car…

What it does

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

Domain
Language, Vision, Multimodal
Task
Language modeling/generation, Question answering, Quantitative reasoning, Translation, Visual question answering, Code generation, Instruction interpretation

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

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
3.8 × 10²⁶ FLOP

Analysis of GPT-4.5's training cluster yields a median estimate of 187M H100-hours of training. The utilization assumptions we used for the Grok 3 estimate (maybe worth revisiting) were 20 to 40% under the H100 FP8 spec of 2000 teraflop/s. This leads to an estimate of 2.7e26 to 5.4e26 FLOP, or a geomean of 3.8e26 Alternatively, using a plausible range of 20 to 50% utilization, given the possibility of FP8 training, yields a median estimate of ~2e25 FLOP. See notebook below for details. https:/…

How it was established
Benchmarks

The training run

What it physically took to train: which chips, how many, for how long, and what that drew from the wall.

Wall-clock time
3,000 hours (125 days)

Likely trained for around 4 months/120 days https://colab.research.google.com/drive/1QBmVPm64Ti0xucN0EsZTgSz_I7Mj9hAZ#scrollTo=NYH1ABJuLJlw

Compute cost
$366,010,628
Cloud vendor
Azure AI
Data centre
Microsoft Goodyear Arizona

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
API access
Training code
Unreleased

How it is classified

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

Frontier model
Yes
Likely above 10²³ FLOP
Yes
Why it is tracked
Training cost

Described by OpenAI as a "new order of magnitude of compute" https://openai.com/index/introducing-gpt-4-5/

Record confidence
Likely

Sources

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

Reference
Introducing GPT-4.5
Last updated
8 June 2026

What the numbers mean

What this model is

GPT-4.5 was published by OpenAI, in United States of America, in February 2025. industry is the category the publisher falls under.

It works in Language, Vision, Multimodal, and is recorded as doing language modeling/generation, Question answering, Quantitative reasoning, Translation, Visual question answering, Code generation, Instruction interpretation.

Because the weights are not available, none of the hardware figures elsewhere on this site apply to it.

What went into building it

Training it took roughly 3.8 × 10²⁶ FLOP of computation — a measure of what producing the model cost, not of how fast it answers.

It is tracked in the underlying dataset for one reason in particular: training cost.

Answers

GPT-4.5 — common questions

01

What GPU do I need to run GPT-4.5?

None. GPT-4.5 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.

02

Is GPT-4.5 open source?

No. GPT-4.5 has not had its weights published, so it exists only as a service controlled by its owner.

03

How many parameters does GPT-4.5 have?

No parameter count has been published for GPT-4.5, which is why no memory or speed figure appears on this page.

04

Who created GPT-4.5?

GPT-4.5 was published by OpenAI, based in United States of America, categorised as industry.

05

When was GPT-4.5 released?

GPT-4.5 was published in February 2025.

06

What is GPT-4.5 used for?

GPT-4.5 works in Language, Vision, Multimodal, and is recorded as handling language modeling/generation, Question answering, Quantitative reasoning, Translation, Visual question answering, Code generation, Instruction interpretation. Models frequently carry more than one of each, and the tags describe purpose rather than capability limits.

07

How much compute was used to train GPT-4.5?

Around 3.8 × 10²⁶ 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.

Source

Original publication

Record last updated 8 June 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.