AraELECTRA TPS calculator

Open weights American University of Beirut 136M parameters March 2021

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 · 271 tok/s

Fastest card

B200

24,913 tok/s · 180 GB

Which GPUs can run AraELECTRA?

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
24,913 tok/s 14,948–39,862 · low confidence	B200 NVIDIA	180 GB	8,000 GB/s	Jan 2024	0.8 GB	Q8_0	Comfortable
24,913 tok/s 14,948–39,862 · low confidence	B300 NVIDIA	288 GB	8,000 GB/s	Sep 2025	0.8 GB	Q8_0	Comfortable
19,894 tok/s 11,936–31,830 · low confidence	Radeon Instinct MI350X AMD	288 GB	8,190 GB/s	Jan 2025	0.8 GB	Q8_0	Comfortable
19,894 tok/s 11,936–31,830 · low confidence	Radeon Instinct MI355X AMD	288 GB	8,190 GB/s	Jan 2025	0.8 GB	Q8_0	Comfortable
15,910 tok/s 9,546–25,457 · low confidence	Radeon Instinct MI300 AMD	128 GB	6,550 GB/s	Jan 2023	0.8 GB	Q8_0	Comfortable
15,228 tok/s 9,137–24,365 · low confidence	H200 NVL NVIDIA	141 GB	4,890 GB/s	Nov 2024	0.8 GB	Q8_0	Comfortable
15,228 tok/s 9,137–24,365 · low confidence	H200 SXM 141 GB NVIDIA	141 GB	4,890 GB/s	Nov 2024	0.8 GB	Q8_0	Comfortable
14,574 tok/s 8,745–23,319 · low confidence	Radeon Instinct MI325X AMD	256 GB	6,000 GB/s	Oct 2024	0.8 GB	Q8_0	Comfortable
12,935 tok/s 7,761–20,696 · low confidence	Radeon Instinct MI300A AMD	128 GB	5,325 GB/s	Dec 2023	0.8 GB	Q8_0	Comfortable
12,935 tok/s 7,761–20,696 · low confidence	Radeon Instinct MI300X AMD	192 GB	5,325 GB/s	Dec 2023	0.8 GB	Q8_0	Comfortable
12,935 tok/s 7,761–20,696 · low confidence	Radeon Instinct MI308X AMD	192 GB	5,325 GB/s	Dec 2023	0.8 GB	Q8_0	Comfortable
12,270 tok/s 7,362–19,632 · low confidence	H100 NVL 94 GB NVIDIA	94 GB	3,940 GB/s	Mar 2023	0.8 GB	Q8_0	Comfortable
10,464 tok/s 6,278–16,742 · low confidence	H100 PCIe 96 GB NVIDIA	96 GB	3,360 GB/s	Mar 2023	0.8 GB	Q8_0	Comfortable
10,464 tok/s 6,278–16,742 · low confidence	H100 SXM5 80 GB NVIDIA	80 GB	3,360 GB/s	Oct 2022	0.8 GB	Q8_0	Comfortable
10,464 tok/s 6,278–16,742 · low confidence	H100 SXM5 94 GB NVIDIA	94 GB	3,360 GB/s	Mar 2023	0.8 GB	Q8_0	Comfortable
10,464 tok/s 6,278–16,742 · low confidence	H100 SXM5 96 GB NVIDIA	96 GB	3,360 GB/s	Mar 2023	0.8 GB	Q8_0	Comfortable
10,464 tok/s 6,278–16,742 · low confidence	H800 SXM5 NVIDIA	80 GB	3,360 GB/s	Mar 2023	0.8 GB	Q8_0	Comfortable
7,967 tok/s 4,780–12,748 · low confidence	Radeon Instinct MI250 AMD	128 GB	3,280 GB/s	Nov 2021	0.8 GB	Q8_0	Comfortable
7,967 tok/s 4,780–12,748 · low confidence	Radeon Instinct MI250X AMD	128 GB	3,280 GB/s	Nov 2021	0.8 GB	Q8_0	Comfortable
6,639 tok/s 3,984–10,623 · low confidence	Data Center GPU Max 1550 Intel	128 GB	3,280 GB/s	Jan 2023	0.8 GB	Q8_0	Comfortable
6,498 tok/s 3,899–10,396 · low confidence	Data Center GPU Max Subsystem Intel	128 GB	3,210 GB/s	Jan 2023	0.8 GB	Q8_0	Comfortable
6,353 tok/s 3,812–10,165 · low confidence	A100 SXM4 80 GB NVIDIA	80 GB	2,040 GB/s	Nov 2020	0.8 GB	Q8_0	Comfortable
6,353 tok/s 3,812–10,165 · low confidence	A100X NVIDIA	80 GB	2,040 GB/s	Jun 2021	0.8 GB	Q8_0	Comfortable
6,353 tok/s 3,812–10,165 · low confidence	A800 SXM4 80 GB NVIDIA	80 GB	2,040 GB/s	Aug 2022	0.8 GB	Q8_0	Comfortable
6,353 tok/s 3,812–10,165 · low confidence	H100 CNX NVIDIA	80 GB	2,040 GB/s	Mar 2023	0.8 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: 7 March 2021
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: 136M
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: 2.6 × 10²⁰ 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: 8
Wall-clock time: 576 hours (24 days)
Power draw: 7.3 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)
Training code: Open source
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: AraELECTRA: Pre-Training Text Discriminators for Arabic Language Understanding
Last updated: 28 November 2025

The extremes

The ten fastest GPUs that run AraELECTRA

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 AraELECTRA

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

What it takes to run this model

Minimum card

Tesla C1080

Memory needed

0.8 GB

Fastest

24,913 tok/s

AraELECTRA is small enough at 136M parameters that hardware is rarely the obstacle — 818 of the cards we track can run it, including cards several years old.

The smallest card that holds it is the Tesla C1080 with 4 GB, running it at Q8_0 and producing around 271 tokens per second.

Top of the range is the B200, at roughly 24,913 tokens per second thanks to 8,000 GB/s of bandwidth.

About this model

AraELECTRA was published by American University of Beirut, in Lebanon, in March 2021. academia is the category the publisher falls under.

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

The weights being open is what puts this page in the calculator rather than only in the catalogue: it is a model you can actually hold. It is published under the aubmindlab organisation on Hugging Face.

How fast it runs, and why

The median result is around 699.6 tokens per second; 818 cards produce text faster than most people read it.

Every weight participates in every token here, so bandwidth is the whole story: the ranking below is effectively a ranking of memory throughput.

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

Producing it required around 2.6 × 10²⁰ FLOP of arithmetic, on Google TPU v3, which is a statement about the training budget rather than about inference.

Step by step

How to choose a GPU for AraELECTRA

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

01

Read the memory figure first

Look at what AraELECTRA actually needs — around 0.8 GB at Q8_0. No amount of processing power compensates for a card that cannot hold it.
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 AraELECTRA.
03

Set a quality floor

Compression is what makes AraELECTRA fit smaller cards, at some cost in accuracy — Q8_0 on the smallest card that fits. A minimum quality removes the ones that go too far.
04

Sort by speed

Ranking by tokens per second for AraELECTRA follows memory bandwidth, not core counts, which is why the B200 tops it at 24,913 tok/s.
05

Look at the headroom, not just the fit

The fit column separates cards that just manage AraELECTRA from those with room to spare. Buy for the second if the context might grow.
06

See what else that card runs

Every card name links to its own page, which runs the same calculation across the whole model catalogue. Worth a look before buying for AraELECTRA alone — a card is usually bought for more than one model.

Answers

AraELECTRA — common questions

How much compute was used to train AraELECTRA?

Around 2.6 × 10²⁰ 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 AraELECTRA if it does not fit in my GPU?

Only by offloading, which is usually a false economy: the part in system memory drags the whole thing down. Our figures for AraELECTRA assume it is fully resident.

Would two GPUs run AraELECTRA faster?

Two cards buy memory rather than speed. That matters for AraELECTRA only if one card cannot hold it — 818 can, so a second adds little.

Why does the quantisation differ between cards for AraELECTRA?

Because capacity varies, so does how hard AraELECTRA has to be squeezed — 1 distinct levels appear in the table above. Set a minimum quality to compare at one.

How accurate are these AraELECTRA speed estimates?

They are calculated from specifications rather than measured, and each carries a range — 14,948–39,862 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 AraELECTRA?

The smallest card in our catalogue that holds AraELECTRA is the Tesla C1080, with 4 GB of memory. It runs the model at Q8_0 using about 0.8 GB, and produces roughly 271 tokens per second. 818 cards in total can run it.

How fast is AraELECTRA on a GPU?

It depends on the card. The quickest we calculate is a B200 at about 24,913 tokens per second; the slowest that still runs it manages considerably less. Reading speed is around ten tokens per second, and 818 of the cards that can run AraELECTRA clear that.

How much VRAM does AraELECTRA need?

About 0.8 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 AraELECTRA on a 8 GB GPU?

Yes. A CMP 170HX 8 GB with 8 GB runs it at Q8_0, using about 0.8 GB and generating roughly 4,640 tokens per second — a comfortable fit.

Can I run AraELECTRA on a 12 GB GPU?

Yes. A GeForce RTX 3080 Ti with 12 GB runs it at Q8_0, using about 0.8 GB and generating roughly 2,841 tokens per second — a comfortable fit.

Can I run AraELECTRA on a 16 GB GPU?

Yes. A Tesla V100 SXM2 16 GB with 16 GB runs it at Q8_0, using about 0.8 GB and generating roughly 3,519 tokens per second — a comfortable fit.

Can I run AraELECTRA on a 24 GB GPU?

Yes. A GeForce RTX 5090 D V2 with 24 GB runs it at Q8_0, using about 0.8 GB and generating roughly 4,173 tokens per second — a comfortable fit.

Is AraELECTRA open source?

Its weights are published, so AraELECTRA 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.

How many parameters does AraELECTRA have?

AraELECTRA has 136M parameters. 12 encoder layers, 12 attention heads, 768 hidden size, and 512 maximum input sequence length for a total of 136M 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 AraELECTRA?

AraELECTRA was published by American University of Beirut, based in Lebanon, categorised as academia.

When was AraELECTRA released?

AraELECTRA was published in March 2021. 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 AraELECTRA used for?

AraELECTRA 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 AraELECTRA?

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.

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