AraBERT LArge v2 TPS calculator

Open weights American University of Beirut 371M parameters March 2020

Each card below is assessed against this model at the context length and minimum quality you choose. Speed is an estimate for a single request, calculated from the card's memory bandwidth and the size of the model once compressed.

Calculated for this model

818 cards that can run it

818 cards we hold specifications for

Smallest card that fits

Tesla C1080

4 GB · Q8_0 · 99.4 tok/s

Fastest card

B200

9,133 tok/s · 180 GB

Which GPUs can run AraBERT LArge v2?

Set the inputs, read the answer

Context length tokens

A longer conversation needs more memory, which can push this model off smaller cards.

Minimum quality

Hides cards that would only fit the model by compressing it below this point.

Search cards

Card memory

Also show cards that cannot run it

818 cards match

Calculating

					Needs	Quantisation	Fit
9,133 tok/s 5,480–14,612 · low confidence	B200 NVIDIA	180 GB	8,000 GB/s	Jan 2024	1.1 GB	Q8_0	Comfortable
9,133 tok/s 5,480–14,612 · low confidence	B300 NVIDIA	288 GB	8,000 GB/s	Sep 2025	1.1 GB	Q8_0	Comfortable
7,293 tok/s 4,376–11,668 · low confidence	Radeon Instinct MI350X AMD	288 GB	8,190 GB/s	Jan 2025	1.1 GB	Q8_0	Comfortable
7,293 tok/s 4,376–11,668 · low confidence	Radeon Instinct MI355X AMD	288 GB	8,190 GB/s	Jan 2025	1.1 GB	Q8_0	Comfortable
5,832 tok/s 3,499–9,332 · low confidence	Radeon Instinct MI300 AMD	128 GB	6,550 GB/s	Jan 2023	1.1 GB	Q8_0	Comfortable
5,582 tok/s 3,349–8,932 · low confidence	H200 NVL NVIDIA	141 GB	4,890 GB/s	Nov 2024	1.1 GB	Q8_0	Comfortable
5,582 tok/s 3,349–8,932 · low confidence	H200 SXM 141 GB NVIDIA	141 GB	4,890 GB/s	Nov 2024	1.1 GB	Q8_0	Comfortable
5,343 tok/s 3,206–8,548 · low confidence	Radeon Instinct MI325X AMD	256 GB	6,000 GB/s	Oct 2024	1.1 GB	Q8_0	Comfortable
4,742 tok/s 2,845–7,587 · low confidence	Radeon Instinct MI300A AMD	128 GB	5,325 GB/s	Dec 2023	1.1 GB	Q8_0	Comfortable
4,742 tok/s 2,845–7,587 · low confidence	Radeon Instinct MI300X AMD	192 GB	5,325 GB/s	Dec 2023	1.1 GB	Q8_0	Comfortable
4,742 tok/s 2,845–7,587 · low confidence	Radeon Instinct MI308X AMD	192 GB	5,325 GB/s	Dec 2023	1.1 GB	Q8_0	Comfortable
4,498 tok/s 2,699–7,197 · low confidence	H100 NVL 94 GB NVIDIA	94 GB	3,940 GB/s	Mar 2023	1.1 GB	Q8_0	Comfortable
3,836 tok/s 2,301–6,137 · low confidence	H100 PCIe 96 GB NVIDIA	96 GB	3,360 GB/s	Mar 2023	1.1 GB	Q8_0	Comfortable
3,836 tok/s 2,301–6,137 · low confidence	H100 SXM5 80 GB NVIDIA	80 GB	3,360 GB/s	Oct 2022	1.1 GB	Q8_0	Comfortable
3,836 tok/s 2,301–6,137 · low confidence	H100 SXM5 94 GB NVIDIA	94 GB	3,360 GB/s	Mar 2023	1.1 GB	Q8_0	Comfortable
3,836 tok/s 2,301–6,137 · low confidence	H100 SXM5 96 GB NVIDIA	96 GB	3,360 GB/s	Mar 2023	1.1 GB	Q8_0	Comfortable
3,836 tok/s 2,301–6,137 · low confidence	H800 SXM5 NVIDIA	80 GB	3,360 GB/s	Mar 2023	1.1 GB	Q8_0	Comfortable
2,921 tok/s 1,752–4,673 · low confidence	Radeon Instinct MI250 AMD	128 GB	3,280 GB/s	Nov 2021	1.1 GB	Q8_0	Comfortable
2,921 tok/s 1,752–4,673 · low confidence	Radeon Instinct MI250X AMD	128 GB	3,280 GB/s	Nov 2021	1.1 GB	Q8_0	Comfortable
2,434 tok/s 1,460–3,894 · low confidence	Data Center GPU Max 1550 Intel	128 GB	3,280 GB/s	Jan 2023	1.1 GB	Q8_0	Comfortable
2,382 tok/s 1,429–3,811 · low confidence	Data Center GPU Max Subsystem Intel	128 GB	3,210 GB/s	Jan 2023	1.1 GB	Q8_0	Comfortable
2,329 tok/s 1,397–3,726 · low confidence	A100 SXM4 80 GB NVIDIA	80 GB	2,040 GB/s	Nov 2020	1.1 GB	Q8_0	Comfortable
2,329 tok/s 1,397–3,726 · low confidence	A100X NVIDIA	80 GB	2,040 GB/s	Jun 2021	1.1 GB	Q8_0	Comfortable
2,329 tok/s 1,397–3,726 · low confidence	A800 SXM4 80 GB NVIDIA	80 GB	2,040 GB/s	Aug 2022	1.1 GB	Q8_0	Comfortable
2,329 tok/s 1,397–3,726 · low confidence	H100 CNX NVIDIA	80 GB	2,040 GB/s	Mar 2023	1.1 GB	Q8_0	Comfortable

Speeds are estimates for a single request — one conversation at a time — calculated from memory bandwidth, model size and quantisation. Real throughput varies with the inference runtime and its version. Figures published by hardware vendors measure many simultaneous requests and are much higher.

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: American University of Beirut
Organisation type: Academia
Country: Lebanon
Published: 30 March 2020
Authors: Wissam Antoun, Fady Baly, Hazem Hajj

What it does

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

Domain: Language
Task: Language modeling/generation, Question answering

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: 371M
Training data: tokens
Epochs: 75

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: 1.5 × 10²¹ FLOP
How it was established: Operation counting,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: Google TPU v3
Chips used: 128
Wall-clock time: 168 hours (7 days)
Power draw: 117.7 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)
Hugging Face: aubmindlab

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: AraBERT v1 & v2 : Pre-training BERT for Arabic Language Understanding
Last updated: 28 November 2025

The extremes

The ten fastest GPUs that run AraBERT LArge v2

Ranked by estimated tokens per second, newest card first where speeds tie. Because generation is bound by memory bandwidth, this ordering follows bandwidth rather than any gaming benchmark.

The smallest GPUs that still run AraBERT LArge v2

The cheapest route in, by memory capacity. A tight fit runs the model but leaves nothing spare for a longer conversation.

What the numbers mean

Hardware requirements in practice

Minimum card

Tesla C1080

Memory needed

1.1 GB

Fastest

9,133 tok/s

AraBERT LArge v2 is small enough at 371M parameters that hardware is rarely the obstacle — 818 of the cards we track can run it, including cards several years old.

The entry point is the Tesla C1080: 4 GB of memory, Q8_0 compression, roughly 99.4 tokens per second.

Top of the range is the B200, at roughly 9,133 tokens per second thanks to 8,000 GB/s of bandwidth.

What this model is

AraBERT LArge v2 was published by American University of Beirut, in Lebanon, in March 2020. academia is the category the publisher falls under.

It works in Language, and is recorded as doing language modeling/generation, Question answering.

Published weights mean the model runs on your machine rather than someone else's, which is what makes the hardware question below answerable at all. It is published under the aubmindlab organisation on Hugging Face.

What decides the speed

Across every card that can run it, the middle of the range is about 256.5 tokens per second, and 817 of them clear the ten tokens per second that roughly matches reading speed.

It is a dense model, so every parameter is read for every token produced. That makes speed track memory bandwidth almost exactly — a card with twice the bandwidth generates roughly twice as fast.

Memory here is estimated from size rather than computed from the architecture, which is not recorded for this model — the numbers are indicative rather than exact.

What went into building it

Training it took roughly 1.5 × 10²¹ FLOP of computation, on Google TPU v3 — a measure of what producing the model cost, not of how fast it answers.

Step by step

How to choose a GPU for AraBERT LArge v2

The table above has already assessed every card we hold specifications for against this model. Getting to your answer takes six steps.

01

Check what it needs before anything else

The table lists every card that can hold AraBERT LArge v2 — around 1.1 GB at Q8_0. That figure, not the card's headline performance, is what decides whether it runs.
02

Match the context to your actual use

Longer conversations cost memory on top of what the weights need. Move the slider to your real working length before trusting any row for AraBERT LArge v2.
03

Decide how much compression you will accept

Each card runs the least-compressed copy it can hold — Q8_0 on the smallest card that fits. Setting a floor drops the cards that only manage AraBERT LArge v2 by squeezing it further than you would want.
04

Sort by speed

Ranking by tokens per second for AraBERT LArge v2 follows memory bandwidth, not core counts, which is why the B200 tops it at 9,133 tok/s.
05

Check the fit verdict before buying

Tight means AraBERT LArge v2 loads and works, with no room to raise the context later. Comfortable means you can. The difference matters more than a few tokens per second.
06

Open the card you have settled on

Each card page repeats this sweep for every model we hold. It answers what else the hardware is good for, beyond AraBERT LArge v2.

Answers

AraBERT LArge v2 — common questions

How many parameters does AraBERT LArge v2 have?

AraBERT LArge v2 has 371M parameters. 371M. 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 AraBERT LArge v2?

AraBERT LArge v2 was published by American University of Beirut, based in Lebanon, categorised as academia.

When was AraBERT LArge v2 released?

AraBERT LArge v2 was published in March 2020. 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 AraBERT LArge v2 used for?

AraBERT LArge v2 works in Language, and is recorded as handling language modeling/generation, Question answering. 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.

Where can I download AraBERT LArge v2?

Its weights are published under the aubmindlab organisation on Hugging Face. We do not host model files — this site calculates what hardware is needed to run them.

How much compute was used to train AraBERT LArge v2?

Around 1.5 × 10²¹ 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.

Can I run AraBERT LArge v2 if it does not fit in my GPU?

Partly. Layers that do not fit sit in system memory and run at a fraction of the speed, so a mostly-offloaded AraBERT LArge v2 is rarely worth using. Every figure here assumes the whole model is on the card.

Would two GPUs run AraBERT LArge v2 faster?

Capacity adds across cards; throughput does not. Since 818 of the cards we track already hold AraBERT LArge v2 on their own, a second card is rarely the answer here.

Why does the quantisation differ between cards for AraBERT LArge v2?

A larger card holds a more accurate copy. Across the cards that run AraBERT LArge v2, 1 compression levels are used; the floor control above pins it to one.

How accurate are these AraBERT LArge v2 speed estimates?

They are calculated from specifications rather than measured, and each carries a range — 5,480–14,612 tok/s on the B200, for instance. The same model and card vary by thirty to fifty per cent depending on the inference software and its version.

What GPU do I need to run AraBERT LArge v2?

The smallest card in our catalogue that holds AraBERT LArge v2 is the Tesla C1080, with 4 GB of memory. It runs the model at Q8_0 using about 1.1 GB, and produces roughly 99.4 tokens per second. 818 cards in total can run it.

How fast is AraBERT LArge v2 on a GPU?

It depends on the card. The quickest we calculate is a B200 at about 9,133 tokens per second; the slowest that still runs it manages considerably less. Reading speed is around ten tokens per second, and 817 of the cards that can run AraBERT LArge v2 clear that.

How much VRAM does AraBERT LArge v2 need?

About 1.1 GB at Q8_0 compression, which is what the smallest card that runs it uses. Less compression needs more: the figures in the memory column above are recalculated for each card, because each one holds the least-compressed version it can.

Can I run AraBERT LArge v2 on a 8 GB GPU?

Yes. A CMP 170HX 8 GB with 8 GB runs it at Q8_0, using about 1.1 GB and generating roughly 1,701 tokens per second — a comfortable fit.

Can I run AraBERT LArge v2 on a 12 GB GPU?

Yes. A GeForce RTX 3080 Ti with 12 GB runs it at Q8_0, using about 1.1 GB and generating roughly 1,042 tokens per second — a comfortable fit.

Can I run AraBERT LArge v2 on a 16 GB GPU?

Yes. A Tesla V100 SXM2 16 GB with 16 GB runs it at Q8_0, using about 1.1 GB and generating roughly 1,290 tokens per second — a comfortable fit.

Can I run AraBERT LArge v2 on a 24 GB GPU?

Yes. A GeForce RTX 5090 D V2 with 24 GB runs it at Q8_0, using about 1.1 GB and generating roughly 1,530 tokens per second — a comfortable fit.

Is AraBERT LArge v2 open source?

Its weights are published, so AraBERT LArge v2 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.

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.

All AI models