Segment Anything Model 2 TPS calculator

Open weights Meta AI 224.4M parameters July 2024

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 · 164 tok/s

Fastest card

B200

15,099 tok/s · 180 GB

Which GPUs can run Segment Anything Model 2?

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
15,099 tok/s 9,059–24,159 · low confidence	B200 NVIDIA	180 GB	8,000 GB/s	Jan 2024	0.9 GB	Q8_0	Comfortable
15,099 tok/s 9,059–24,159 · low confidence	B300 NVIDIA	288 GB	8,000 GB/s	Sep 2025	0.9 GB	Q8_0	Comfortable
12,057 tok/s 7,234–19,291 · low confidence	Radeon Instinct MI350X AMD	288 GB	8,190 GB/s	Jan 2025	0.9 GB	Q8_0	Comfortable
12,057 tok/s 7,234–19,291 · low confidence	Radeon Instinct MI355X AMD	288 GB	8,190 GB/s	Jan 2025	0.9 GB	Q8_0	Comfortable
9,643 tok/s 5,786–15,428 · low confidence	Radeon Instinct MI300 AMD	128 GB	6,550 GB/s	Jan 2023	0.9 GB	Q8_0	Comfortable
9,229 tok/s 5,538–14,767 · low confidence	H200 NVL NVIDIA	141 GB	4,890 GB/s	Nov 2024	0.9 GB	Q8_0	Comfortable
9,229 tok/s 5,538–14,767 · low confidence	H200 SXM 141 GB NVIDIA	141 GB	4,890 GB/s	Nov 2024	0.9 GB	Q8_0	Comfortable
8,833 tok/s 5,300–14,133 · low confidence	Radeon Instinct MI325X AMD	256 GB	6,000 GB/s	Oct 2024	0.9 GB	Q8_0	Comfortable
7,839 tok/s 4,704–12,543 · low confidence	Radeon Instinct MI300A AMD	128 GB	5,325 GB/s	Dec 2023	0.9 GB	Q8_0	Comfortable
7,839 tok/s 4,704–12,543 · low confidence	Radeon Instinct MI300X AMD	192 GB	5,325 GB/s	Dec 2023	0.9 GB	Q8_0	Comfortable
7,839 tok/s 4,704–12,543 · low confidence	Radeon Instinct MI308X AMD	192 GB	5,325 GB/s	Dec 2023	0.9 GB	Q8_0	Comfortable
7,436 tok/s 4,462–11,898 · low confidence	H100 NVL 94 GB NVIDIA	94 GB	3,940 GB/s	Mar 2023	0.9 GB	Q8_0	Comfortable
6,342 tok/s 3,805–10,147 · low confidence	H100 PCIe 96 GB NVIDIA	96 GB	3,360 GB/s	Mar 2023	0.9 GB	Q8_0	Comfortable
6,342 tok/s 3,805–10,147 · low confidence	H100 SXM5 80 GB NVIDIA	80 GB	3,360 GB/s	Oct 2022	0.9 GB	Q8_0	Comfortable
6,342 tok/s 3,805–10,147 · low confidence	H100 SXM5 94 GB NVIDIA	94 GB	3,360 GB/s	Mar 2023	0.9 GB	Q8_0	Comfortable
6,342 tok/s 3,805–10,147 · low confidence	H100 SXM5 96 GB NVIDIA	96 GB	3,360 GB/s	Mar 2023	0.9 GB	Q8_0	Comfortable
6,342 tok/s 3,805–10,147 · low confidence	H800 SXM5 NVIDIA	80 GB	3,360 GB/s	Mar 2023	0.9 GB	Q8_0	Comfortable
4,829 tok/s 2,897–7,726 · low confidence	Radeon Instinct MI250 AMD	128 GB	3,280 GB/s	Nov 2021	0.9 GB	Q8_0	Comfortable
4,829 tok/s 2,897–7,726 · low confidence	Radeon Instinct MI250X AMD	128 GB	3,280 GB/s	Nov 2021	0.9 GB	Q8_0	Comfortable
4,024 tok/s 2,414–6,438 · low confidence	Data Center GPU Max 1550 Intel	128 GB	3,280 GB/s	Jan 2023	0.9 GB	Q8_0	Comfortable
3,938 tok/s 2,363–6,301 · low confidence	Data Center GPU Max Subsystem Intel	128 GB	3,210 GB/s	Jan 2023	0.9 GB	Q8_0	Comfortable
3,850 tok/s 2,310–6,160 · low confidence	A100 SXM4 80 GB NVIDIA	80 GB	2,040 GB/s	Nov 2020	0.9 GB	Q8_0	Comfortable
3,850 tok/s 2,310–6,160 · low confidence	A100X NVIDIA	80 GB	2,040 GB/s	Jun 2021	0.9 GB	Q8_0	Comfortable
3,850 tok/s 2,310–6,160 · low confidence	A800 SXM4 80 GB NVIDIA	80 GB	2,040 GB/s	Aug 2022	0.9 GB	Q8_0	Comfortable
3,850 tok/s 2,310–6,160 · low confidence	H100 CNX NVIDIA	80 GB	2,040 GB/s	Mar 2023	0.9 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: Meta AI
Organisation type: Industry
Country: United States of America
Published: 29 July 2024
Authors: Nikhila Ravi, Valentin Gabeur, Yuan-Ting Hu, Ronghang Hu, Chaitanya Ryali, Tengyu Ma, Haitham Khedr, Roman Rädle, Chloe Rolland, Laura Gustafson, Eric Mintun, Junting Pan, Kalyan Vasudev Alwala, Nicolas Carion, Chao-Yuan Wu, Ross Girshick, Piotr Dollár, Christoph Feichtenhofer

What it does

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

Domain: Vision, Video
Task: Image segmentation

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: 224.4M
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: 1.2 × 10²² FLOP

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: facebook

How it is classified

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

Record confidence: Likely

Sources

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

Reference: SAM 2: Segment Anything in Images and Videos
Last updated: 28 November 2025

The extremes

The ten fastest GPUs that run Segment Anything Model 2

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 Segment Anything Model 2

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

0.9 GB

Fastest

15,099 tok/s

Segment Anything Model 2 is small enough at 224.4M 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 164 tokens per second.

The quickest result comes from a B200 at around 15,099 tokens per second — its 8,000 GB/s of bandwidth is what buys that.

Background

Segment Anything Model 2 was published by Meta AI, in United States of America, in July 2024. The organisation is categorised as industry.

It works in Vision, Video, and is recorded as doing image segmentation.

The weights are published, so it can be downloaded and run on your own hardware indefinitely, offline, with no account attached. It is published under the facebook organisation on Hugging Face.

Reading the throughput figures

Half the cards that hold it manage more than 424.0 tokens per second, and 818 exceed reading speed outright.

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.

How it was trained

Producing it required around 1.2 × 10²² FLOP of arithmetic, which is a statement about the training budget rather than about inference.

Step by step

How to choose a GPU for Segment Anything Model 2

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

01

Start from the memory column

Every card here has been checked against Segment Anything Model 2 — around 0.9 GB at Q8_0. Capacity is the gate — a card either holds it or it does not.
02

Match the context to your actual use

Set the context to what you will actually use. The cache grows with the conversation, and it is the usual reason Segment Anything Model 2 stops fitting a card that seemed fine.
03

Decide how much compression you will accept

Compression is what makes Segment Anything Model 2 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

Compare tokens per second, not specifications

The speed ordering for Segment Anything Model 2 is effectively an ordering by memory bandwidth, which is why the B200 tops it at 15,099 tok/s.
05

Look at the headroom, not just the fit

Tight means Segment Anything Model 2 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

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 Segment Anything Model 2 alone — a card is usually bought for more than one model.

Answers

Segment Anything Model 2 — common questions

Can I run Segment Anything Model 2 on a 8 GB GPU?

Yes. A CMP 170HX 8 GB with 8 GB runs it at Q8_0, using about 0.9 GB and generating roughly 2,812 tokens per second — a comfortable fit.

Can I run Segment Anything Model 2 on a 12 GB GPU?

Yes. A GeForce RTX 3080 Ti with 12 GB runs it at Q8_0, using about 0.9 GB and generating roughly 1,722 tokens per second — a comfortable fit.

Can I run Segment Anything Model 2 on a 16 GB GPU?

Yes. A Tesla V100 SXM2 16 GB with 16 GB runs it at Q8_0, using about 0.9 GB and generating roughly 2,133 tokens per second — a comfortable fit.

Can I run Segment Anything Model 2 on a 24 GB GPU?

Yes. A GeForce RTX 5090 D V2 with 24 GB runs it at Q8_0, using about 0.9 GB and generating roughly 2,529 tokens per second — a comfortable fit.

Is Segment Anything Model 2 open source?

Its weights are published, so Segment Anything Model 2 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 Segment Anything Model 2 have?

Segment Anything Model 2 has 224.4M parameters. sam2.1_hiera_large, 224.4M, https://github.com/facebookresearch/sam2. 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 Segment Anything Model 2?

Segment Anything Model 2 was published by Meta AI, based in United States of America, categorised as industry.

When was Segment Anything Model 2 released?

Segment Anything Model 2 was published in July 2024. 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 Segment Anything Model 2 used for?

Segment Anything Model 2 works in Vision, Video, and is recorded as handling image segmentation. Models frequently carry more than one of each, and the tags describe purpose rather than capability limits.

Where can I download Segment Anything Model 2?

Its weights are published under the facebook 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 Segment Anything Model 2?

Around 1.2 × 10²² 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.

Can I run Segment Anything Model 2 if it does not fit in my GPU?

It can be split between the card and system memory, but Segment Anything Model 2 generates painfully slowly that way. Nothing on this page assumes offloading.

Would two GPUs run Segment Anything Model 2 faster?

Capacity adds across cards; throughput does not. Since 818 of the cards we track already hold Segment Anything Model 2 on their own, a second card is rarely the answer here.

Why does the quantisation differ between cards for Segment Anything Model 2?

Because capacity varies, so does how hard Segment Anything Model 2 has to be squeezed — 1 distinct levels appear in the table above. Set a minimum quality to compare at one.

How accurate are these Segment Anything Model 2 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 9,059–24,159 tok/s on the B200 rather than a single number.

What GPU do I need to run Segment Anything Model 2?

The smallest card in our catalogue that holds Segment Anything Model 2 is the Tesla C1080, with 4 GB of memory. It runs the model at Q8_0 using about 0.9 GB, and produces roughly 164 tokens per second. 818 cards in total can run it.

How fast is Segment Anything Model 2 on a GPU?

It depends on the card. The quickest we calculate is a B200 at about 15,099 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 Segment Anything Model 2 clear that.

How much VRAM does Segment Anything Model 2 need?

About 0.9 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.

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