CLIP (ViT L/14@336px) TPS calculator

Open weights OpenAI 370M parameters January 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 · 99.6 tok/s

Fastest card

B200

9,157 tok/s · 180 GB

Which GPUs can run CLIP (ViT L/14@336px)?

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
9,157 tok/s 5,494–14,652 · low confidence	B200 NVIDIA	180 GB	8,000 GB/s	Jan 2024	1.1 GB	Q8_0	Comfortable
9,157 tok/s 5,494–14,652 · low confidence	B300 NVIDIA	288 GB	8,000 GB/s	Sep 2025	1.1 GB	Q8_0	Comfortable
7,312 tok/s 4,387–11,700 · low confidence	Radeon Instinct MI350X AMD	288 GB	8,190 GB/s	Jan 2025	1.1 GB	Q8_0	Comfortable
7,312 tok/s 4,387–11,700 · low confidence	Radeon Instinct MI355X AMD	288 GB	8,190 GB/s	Jan 2025	1.1 GB	Q8_0	Comfortable
5,848 tok/s 3,509–9,357 · low confidence	Radeon Instinct MI300 AMD	128 GB	6,550 GB/s	Jan 2023	1.1 GB	Q8_0	Comfortable
5,597 tok/s 3,358–8,956 · low confidence	H200 NVL NVIDIA	141 GB	4,890 GB/s	Nov 2024	1.1 GB	Q8_0	Comfortable
5,597 tok/s 3,358–8,956 · low confidence	H200 SXM 141 GB NVIDIA	141 GB	4,890 GB/s	Nov 2024	1.1 GB	Q8_0	Comfortable
5,357 tok/s 3,214–8,571 · low confidence	Radeon Instinct MI325X AMD	256 GB	6,000 GB/s	Oct 2024	1.1 GB	Q8_0	Comfortable
4,754 tok/s 2,853–7,607 · low confidence	Radeon Instinct MI300A AMD	128 GB	5,325 GB/s	Dec 2023	1.1 GB	Q8_0	Comfortable
4,754 tok/s 2,853–7,607 · low confidence	Radeon Instinct MI300X AMD	192 GB	5,325 GB/s	Dec 2023	1.1 GB	Q8_0	Comfortable
4,754 tok/s 2,853–7,607 · low confidence	Radeon Instinct MI308X AMD	192 GB	5,325 GB/s	Dec 2023	1.1 GB	Q8_0	Comfortable
4,510 tok/s 2,706–7,216 · low confidence	H100 NVL 94 GB NVIDIA	94 GB	3,940 GB/s	Mar 2023	1.1 GB	Q8_0	Comfortable
3,846 tok/s 2,308–6,154 · low confidence	H100 PCIe 96 GB NVIDIA	96 GB	3,360 GB/s	Mar 2023	1.1 GB	Q8_0	Comfortable
3,846 tok/s 2,308–6,154 · low confidence	H100 SXM5 80 GB NVIDIA	80 GB	3,360 GB/s	Oct 2022	1.1 GB	Q8_0	Comfortable
3,846 tok/s 2,308–6,154 · low confidence	H100 SXM5 94 GB NVIDIA	94 GB	3,360 GB/s	Mar 2023	1.1 GB	Q8_0	Comfortable
3,846 tok/s 2,308–6,154 · low confidence	H100 SXM5 96 GB NVIDIA	96 GB	3,360 GB/s	Mar 2023	1.1 GB	Q8_0	Comfortable
3,846 tok/s 2,308–6,154 · low confidence	H800 SXM5 NVIDIA	80 GB	3,360 GB/s	Mar 2023	1.1 GB	Q8_0	Comfortable
2,929 tok/s 1,757–4,686 · low confidence	Radeon Instinct MI250 AMD	128 GB	3,280 GB/s	Nov 2021	1.1 GB	Q8_0	Comfortable
2,929 tok/s 1,757–4,686 · low confidence	Radeon Instinct MI250X AMD	128 GB	3,280 GB/s	Nov 2021	1.1 GB	Q8_0	Comfortable
2,440 tok/s 1,464–3,905 · low confidence	Data Center GPU Max 1550 Intel	128 GB	3,280 GB/s	Jan 2023	1.1 GB	Q8_0	Comfortable
2,388 tok/s 1,433–3,821 · low confidence	Data Center GPU Max Subsystem Intel	128 GB	3,210 GB/s	Jan 2023	1.1 GB	Q8_0	Comfortable
2,335 tok/s 1,401–3,736 · low confidence	A100 SXM4 80 GB NVIDIA	80 GB	2,040 GB/s	Nov 2020	1.1 GB	Q8_0	Comfortable
2,335 tok/s 1,401–3,736 · low confidence	A100X NVIDIA	80 GB	2,040 GB/s	Jun 2021	1.1 GB	Q8_0	Comfortable
2,335 tok/s 1,401–3,736 · low confidence	A800 SXM4 80 GB NVIDIA	80 GB	2,040 GB/s	Aug 2022	1.1 GB	Q8_0	Comfortable
2,335 tok/s 1,401–3,736 · 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: OpenAI
Organisation type: Industry
Country: United States of America
Published: 5 January 2021
Authors: Alec Radford, Jong Wook Kim, Chris Hallacy, Aditya Ramesh, Gabriel Goh, Sandhini Agarwal, Girish Sastry, Amanda Askell, Pamela Mishkin, Jack Clark, Gretchen Krueger, Ilya Sutskever

What it does

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

Domain: Multimodal, Vision, Language, Video
Task: Zero-shot image classification, Character recognition (OCR), Video description
Approach: Self-supervised learning
Numerical format: FP16

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: 370M
Training data: 400,000,000 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: 1 × 10²² FLOP
How it was established: Third-party estimation

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: NVIDIA V100
Chips used: 256
Wall-clock time: 288 hours (12 days)
Power draw: 155.9 kW
Compute cost: $24,639

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: 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: Highly cited,SOTA improvement
Record confidence: Confident
Citations: 48,743

Sources

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

Reference: Learning Transferable Visual Models From Natural Language Supervision
Last updated: 25 May 2026

The extremes

The ten fastest GPUs that run CLIP (ViT L/14@336px)

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 CLIP (ViT L/14@336px)

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

1.1 GB

Fastest

9,157 tok/s

CLIP (ViT L/14@336px) is small enough at 370M 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 99.6 tokens per second.

A B200 is the fastest we calculate for it: about 9,157 tokens per second, from 8,000 GB/s of memory bandwidth.

About this model

CLIP (ViT L/14@336px) was published by OpenAI, in United States of America, in January 2021. industry is the category the publisher falls under.

It works in Multimodal, Vision, Language, Video, and is recorded as doing zero-shot image classification, Character recognition (OCR), Video description.

The weights are published, so it can be downloaded and run on your own hardware indefinitely, offline, with no account attached.

How fast it runs, and why

Across every card that can run it, the middle of the range is about 257.1 tokens per second, and 817 of them clear the ten tokens per second that roughly matches reading speed.

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.

Training and provenance

Producing it required around 1 × 10²² FLOP of arithmetic, on NVIDIA V100, which is a statement about the training budget rather than about inference.

Around 400,000,000 tokens went into training it.

It is tracked in the underlying dataset for one reason in particular: highly cited,SOTA improvement.

Step by step

How to choose a GPU for CLIP (ViT L/14@336px)

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

Every card here has been checked against CLIP (ViT L/14@336px) — around 1.1 GB at Q8_0. Capacity is the gate — a card either holds it or it does not.
02

Decide how long your conversations run

Set the context to what you will actually use. The cache grows with the conversation, and it is the usual reason CLIP (ViT L/14@336px) stops fitting a card that seemed fine.
03

Set a quality floor

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 CLIP (ViT L/14@336px) by squeezing it further than you would want.
04

Compare tokens per second, not specifications

Ranking by tokens per second for CLIP (ViT L/14@336px) follows memory bandwidth, not core counts, which is why the B200 tops it at 9,157 tok/s.
05

Look at the headroom, not just the fit

A tight fit runs CLIP (ViT L/14@336px) but leaves nothing spare for a longer conversation; comfortable has headroom. If you expect to grow the context, buy for comfortable.
06

Check the card from the other side

Every card name links to its own page, which runs the same calculation across the whole model catalogue. Worth a look before buying for CLIP (ViT L/14@336px) alone — a card is usually bought for more than one model.

Answers

CLIP (ViT L/14@336px) — common questions

How many parameters does CLIP (ViT L/14@336px) have?

CLIP (ViT L/14@336px) has 370M parameters. Image encoder Vision Transformer Table 1 in https://arxiv.org/pdf/2010.11929.pdf Authors fine-tuned ViT L/14 at additional 336px resolution, hence the @336 (See ViT) 307M params Text encoder ~Transformer (from paper) 63M params. 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 CLIP (ViT L/14@336px)?

CLIP (ViT L/14@336px) was published by OpenAI, based in United States of America, categorised as industry.

When was CLIP (ViT L/14@336px) released?

CLIP (ViT L/14@336px) was published in January 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 CLIP (ViT L/14@336px) used for?

CLIP (ViT L/14@336px) works in Multimodal, Vision, Language, Video, and is recorded as handling zero-shot image classification, Character recognition (OCR), Video description. Models frequently carry more than one of each, and the tags describe purpose rather than capability limits.

Where can I download CLIP (ViT L/14@336px)?

The weights for CLIP (ViT L/14@336px) are published, though we do not hold a repository link for it. This site calculates hardware requirements rather than hosting model files.

How much compute was used to train CLIP (ViT L/14@336px)?

Around 1 × 10²² FLOP, on NVIDIA V100. 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 CLIP (ViT L/14@336px) 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 CLIP (ViT L/14@336px) assume it is fully resident.

Would two GPUs run CLIP (ViT L/14@336px) faster?

A second card roughly doubles the memory available but not the generation rate. With 818 cards already able to run CLIP (ViT L/14@336px) alone, the case for pairing is weak.

Why does the quantisation differ between cards for CLIP (ViT L/14@336px)?

Because capacity varies, so does how hard CLIP (ViT L/14@336px) has to be squeezed — 1 distinct levels appear in the table above. Set a minimum quality to compare at one.

How accurate are these CLIP (ViT L/14@336px) speed estimates?

Every figure is derived from memory bandwidth and model size, not benchmarked. That is why each is published as a range such as 5,494–14,652 tok/s on the B200 rather than a single number.

What GPU do I need to run CLIP (ViT L/14@336px)?

The smallest card in our catalogue that holds CLIP (ViT L/14@336px) is the Tesla C1080, with 4 GB of memory. It runs the model at Q8_0 using about 1.1 GB, and produces roughly 99.6 tokens per second. 818 cards in total can run it.

How fast is CLIP (ViT L/14@336px) on a GPU?

It depends on the card. The quickest we calculate is a B200 at about 9,157 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 CLIP (ViT L/14@336px) clear that.

How much VRAM does CLIP (ViT L/14@336px) 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 CLIP (ViT L/14@336px) 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 1,706 tokens per second — a comfortable fit.

Can I run CLIP (ViT L/14@336px) 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 1,044 tokens per second — a comfortable fit.

Can I run CLIP (ViT L/14@336px) 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 1,293 tokens per second — a comfortable fit.

Can I run CLIP (ViT L/14@336px) 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 1,534 tokens per second — a comfortable fit.

Is CLIP (ViT L/14@336px) open source?

Its weights are published, so CLIP (ViT L/14@336px) 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 25 May 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.

All AI models