CLIP (ResNet-50) TPS calculator

Open weights OpenAI 88.6M 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 · 416 tok/s

Fastest card

B200

38,242 tok/s · 180 GB

Which GPUs can run CLIP (ResNet-50)?

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
38,242 tok/s 22,945–61,187 · low confidence	B200 NVIDIA	180 GB	8,000 GB/s	Jan 2024	0.8 GB	Q8_0	Comfortable
38,242 tok/s 22,945–61,187 · low confidence	B300 NVIDIA	288 GB	8,000 GB/s	Sep 2025	0.8 GB	Q8_0	Comfortable
30,537 tok/s 18,322–48,859 · low confidence	Radeon Instinct MI350X AMD	288 GB	8,190 GB/s	Jan 2025	0.8 GB	Q8_0	Comfortable
30,537 tok/s 18,322–48,859 · low confidence	Radeon Instinct MI355X AMD	288 GB	8,190 GB/s	Jan 2025	0.8 GB	Q8_0	Comfortable
24,422 tok/s 14,653–39,076 · low confidence	Radeon Instinct MI300 AMD	128 GB	6,550 GB/s	Jan 2023	0.8 GB	Q8_0	Comfortable
23,375 tok/s 14,025–37,401 · low confidence	H200 NVL NVIDIA	141 GB	4,890 GB/s	Nov 2024	0.8 GB	Q8_0	Comfortable
23,375 tok/s 14,025–37,401 · low confidence	H200 SXM 141 GB NVIDIA	141 GB	4,890 GB/s	Nov 2024	0.8 GB	Q8_0	Comfortable
22,372 tok/s 13,423–35,794 · low confidence	Radeon Instinct MI325X AMD	256 GB	6,000 GB/s	Oct 2024	0.8 GB	Q8_0	Comfortable
19,855 tok/s 11,913–31,768 · low confidence	Radeon Instinct MI300A AMD	128 GB	5,325 GB/s	Dec 2023	0.8 GB	Q8_0	Comfortable
19,855 tok/s 11,913–31,768 · low confidence	Radeon Instinct MI300X AMD	192 GB	5,325 GB/s	Dec 2023	0.8 GB	Q8_0	Comfortable
19,855 tok/s 11,913–31,768 · low confidence	Radeon Instinct MI308X AMD	192 GB	5,325 GB/s	Dec 2023	0.8 GB	Q8_0	Comfortable
18,834 tok/s 11,300–30,135 · low confidence	H100 NVL 94 GB NVIDIA	94 GB	3,940 GB/s	Mar 2023	0.8 GB	Q8_0	Comfortable
16,062 tok/s 9,637–25,699 · low confidence	H100 PCIe 96 GB NVIDIA	96 GB	3,360 GB/s	Mar 2023	0.8 GB	Q8_0	Comfortable
16,062 tok/s 9,637–25,699 · low confidence	H100 SXM5 80 GB NVIDIA	80 GB	3,360 GB/s	Oct 2022	0.8 GB	Q8_0	Comfortable
16,062 tok/s 9,637–25,699 · low confidence	H100 SXM5 94 GB NVIDIA	94 GB	3,360 GB/s	Mar 2023	0.8 GB	Q8_0	Comfortable
16,062 tok/s 9,637–25,699 · low confidence	H100 SXM5 96 GB NVIDIA	96 GB	3,360 GB/s	Mar 2023	0.8 GB	Q8_0	Comfortable
16,062 tok/s 9,637–25,699 · low confidence	H800 SXM5 NVIDIA	80 GB	3,360 GB/s	Mar 2023	0.8 GB	Q8_0	Comfortable
12,230 tok/s 7,338–19,568 · low confidence	Radeon Instinct MI250 AMD	128 GB	3,280 GB/s	Nov 2021	0.8 GB	Q8_0	Comfortable
12,230 tok/s 7,338–19,568 · low confidence	Radeon Instinct MI250X AMD	128 GB	3,280 GB/s	Nov 2021	0.8 GB	Q8_0	Comfortable
10,191 tok/s 6,115–16,306 · low confidence	Data Center GPU Max 1550 Intel	128 GB	3,280 GB/s	Jan 2023	0.8 GB	Q8_0	Comfortable
9,974 tok/s 5,984–15,958 · low confidence	Data Center GPU Max Subsystem Intel	128 GB	3,210 GB/s	Jan 2023	0.8 GB	Q8_0	Comfortable
9,752 tok/s 5,851–15,603 · low confidence	A100 SXM4 80 GB NVIDIA	80 GB	2,040 GB/s	Nov 2020	0.8 GB	Q8_0	Comfortable
9,752 tok/s 5,851–15,603 · low confidence	A100X NVIDIA	80 GB	2,040 GB/s	Jun 2021	0.8 GB	Q8_0	Comfortable
9,752 tok/s 5,851–15,603 · low confidence	A800 SXM4 80 GB NVIDIA	80 GB	2,040 GB/s	Aug 2022	0.8 GB	Q8_0	Comfortable
9,752 tok/s 5,851–15,603 · 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: 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
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: 88.6M
Training data: 400,000,000 tokens

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: Likely
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 (ResNet-50)

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 (ResNet-50)

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

38,242 tok/s

CLIP (ResNet-50) is small enough at 88.6M 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 416 tokens per second.

The quickest result comes from a B200 at around 38,242 tokens per second — its 8,000 GB/s of bandwidth is what buys that.

About this model

CLIP (ResNet-50) 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 being open is what puts this page in the calculator rather than only in the catalogue: it is a model you can actually hold.

How fast it runs, and why

Across every card that can run it, the middle of the range is about 1,073.8 tokens per second, and 818 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.

Without the attention layout on record, the memory column is an approximation. It is close enough to choose hardware by, and least reliable at long context.

What went into building it

The training set ran to roughly 400,000,000 tokens.

The reason it appears in this catalogue at all is highly cited,SOTA improvement.

Step by step

How to choose a GPU for CLIP (ResNet-50)

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

Every card here has been checked against CLIP (ResNet-50) — around 0.8 GB at Q8_0. Capacity is the gate — a card either holds it or it does not.
02

Set the context length you will work at

The conversation occupies memory too, and grows as it goes. Set the slider to the length you expect: at long context CLIP (ResNet-50) can slip off a card that handles short questions easily.
03

Decide how much compression you will accept

The quantisation column varies by card, because a bigger card holds a more accurate copy of CLIP (ResNet-50) — Q8_0 on the smallest card that fits. Set a floor to hold the comparison at one level.
04

Rank by throughput rather than spec sheet

Ranking by tokens per second for CLIP (ResNet-50) follows memory bandwidth, not core counts, which is why the B200 tops it at 38,242 tok/s.
05

Check the fit verdict before buying

A tight fit runs CLIP (ResNet-50) 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 (ResNet-50) alone — a card is usually bought for more than one model.

Answers

CLIP (ResNet-50) — common questions

Can I run CLIP (ResNet-50) if it does not fit in my GPU?

It can be split between the card and system memory, but CLIP (ResNet-50) generates painfully slowly that way. Nothing on this page assumes offloading.

Would two GPUs run CLIP (ResNet-50) faster?

A second card roughly doubles the memory available but not the generation rate. With 818 cards already able to run CLIP (ResNet-50) alone, the case for pairing is weak.

Why does the quantisation differ between cards for CLIP (ResNet-50)?

Each card is shown running the least-compressed copy it can hold, and CLIP (ResNet-50) appears at 1 different compression levels across the cards that fit it. Bigger cards get the more accurate version.

How accurate are these CLIP (ResNet-50) 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 22,945–61,187 tok/s on the B200 rather than a single number.

What GPU do I need to run CLIP (ResNet-50)?

The smallest card in our catalogue that holds CLIP (ResNet-50) is the Tesla C1080, with 4 GB of memory. It runs the model at Q8_0 using about 0.8 GB, and produces roughly 416 tokens per second. 818 cards in total can run it.

How fast is CLIP (ResNet-50) on a GPU?

It depends on the card. The quickest we calculate is a B200 at about 38,242 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 CLIP (ResNet-50) clear that.

How much VRAM does CLIP (ResNet-50) 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 CLIP (ResNet-50) 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 7,123 tokens per second — a comfortable fit.

Can I run CLIP (ResNet-50) 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 4,361 tokens per second — a comfortable fit.

Can I run CLIP (ResNet-50) 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 5,402 tokens per second — a comfortable fit.

Can I run CLIP (ResNet-50) 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 6,406 tokens per second — a comfortable fit.

Is CLIP (ResNet-50) open source?

Its weights are published, so CLIP (ResNet-50) 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 CLIP (ResNet-50) have?

CLIP (ResNet-50) has 88.6M parameters. Image encoder ~ResNet-50 (from paper) 25.6M 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 (ResNet-50)?

CLIP (ResNet-50) was published by OpenAI, based in United States of America, categorised as industry.

When was CLIP (ResNet-50) released?

CLIP (ResNet-50) 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 (ResNet-50) used for?

CLIP (ResNet-50) works in Multimodal, Vision, Language, Video, and is recorded as handling zero-shot image classification, Character recognition (OCR), Video description. 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 CLIP (ResNet-50)?

The weights for CLIP (ResNet-50) are published, though we do not hold a repository link for it. This site calculates hardware requirements rather than hosting model files.

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