Pixtral 12B TPS calculator

Open weights Mistral AI 12.4B parameters September 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

509 cards that can run it

818 cards we hold specifications for

Smallest card that fits

Xeon Phi 5110P

8 GB · Q3_K_M · 19.2 tok/s

Fastest card

B200

273 tok/s · 180 GB

Which GPUs can run Pixtral 12B?

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

509 cards match

Calculating

					Needs	Quantisation	Fit
273 tok/s 164–437 · low confidence	B200 NVIDIA	180 GB	8,000 GB/s	Jan 2024	14.0 GB	Q8_0	Comfortable
273 tok/s 164–437 · low confidence	B300 NVIDIA	288 GB	8,000 GB/s	Sep 2025	14.0 GB	Q8_0	Comfortable
218 tok/s 131–349 · low confidence	Radeon Instinct MI350X AMD	288 GB	8,190 GB/s	Jan 2025	14.0 GB	Q8_0	Comfortable
218 tok/s 131–349 · low confidence	Radeon Instinct MI355X AMD	288 GB	8,190 GB/s	Jan 2025	14.0 GB	Q8_0	Comfortable
175 tok/s 105–279 · low confidence	Radeon Instinct MI300 AMD	128 GB	6,550 GB/s	Jan 2023	14.0 GB	Q8_0	Comfortable
167 tok/s 100–267 · low confidence	H200 NVL NVIDIA	141 GB	4,890 GB/s	Nov 2024	14.0 GB	Q8_0	Comfortable
167 tok/s 100–267 · low confidence	H200 SXM 141 GB NVIDIA	141 GB	4,890 GB/s	Nov 2024	14.0 GB	Q8_0	Comfortable
160 tok/s 96–256 · low confidence	Radeon Instinct MI325X AMD	256 GB	6,000 GB/s	Oct 2024	14.0 GB	Q8_0	Comfortable
142 tok/s 85–227 · low confidence	Radeon Instinct MI300A AMD	128 GB	5,325 GB/s	Dec 2023	14.0 GB	Q8_0	Comfortable
142 tok/s 85–227 · low confidence	Radeon Instinct MI300X AMD	192 GB	5,325 GB/s	Dec 2023	14.0 GB	Q8_0	Comfortable
142 tok/s 85–227 · low confidence	Radeon Instinct MI308X AMD	192 GB	5,325 GB/s	Dec 2023	14.0 GB	Q8_0	Comfortable
137 tok/s 82–220 · low confidence	CMP 170HX 8 GB NVIDIA	8 GB	1,490 GB/s	Sep 2021	6.8 GB	Q3_K_M	Tight
135 tok/s 81–215 · low confidence	H100 NVL 94 GB NVIDIA	94 GB	3,940 GB/s	Mar 2023	14.0 GB	Q8_0	Comfortable
123 tok/s 74–197 · low confidence	CMP 170HX 10 GB NVIDIA	10 GB	1,560 GB/s	Sep 2021	8.2 GB	Q4_K_M	Tight
115 tok/s 69–184 · low confidence	H100 PCIe 96 GB NVIDIA	96 GB	3,360 GB/s	Mar 2023	14.0 GB	Q8_0	Comfortable
115 tok/s 69–184 · low confidence	H100 SXM5 80 GB NVIDIA	80 GB	3,360 GB/s	Oct 2022	14.0 GB	Q8_0	Comfortable
115 tok/s 69–184 · low confidence	H100 SXM5 94 GB NVIDIA	94 GB	3,360 GB/s	Mar 2023	14.0 GB	Q8_0	Comfortable
115 tok/s 69–184 · low confidence	H100 SXM5 96 GB NVIDIA	96 GB	3,360 GB/s	Mar 2023	14.0 GB	Q8_0	Comfortable
115 tok/s 69–184 · low confidence	H800 SXM5 NVIDIA	80 GB	3,360 GB/s	Mar 2023	14.0 GB	Q8_0	Comfortable
87.4 tok/s 52–140 · low confidence	Radeon Instinct MI250 AMD	128 GB	3,280 GB/s	Nov 2021	14.0 GB	Q8_0	Comfortable
87.4 tok/s 52–140 · low confidence	Radeon Instinct MI250X AMD	128 GB	3,280 GB/s	Nov 2021	14.0 GB	Q8_0	Comfortable
72.8 tok/s 44–117 · low confidence	Data Center GPU Max 1550 Intel	128 GB	3,280 GB/s	Jan 2023	14.0 GB	Q8_0	Comfortable
71.3 tok/s 43–114 · low confidence	Data Center GPU Max Subsystem Intel	128 GB	3,210 GB/s	Jan 2023	14.0 GB	Q8_0	Comfortable
70.8 tok/s 42–113 · low confidence	RTX A5000-8Q NVIDIA	8 GB	768 GB/s	Apr 2021	6.8 GB	Q3_K_M	Tight
69.7 tok/s 42–111 · low confidence	A100 SXM4 80 GB NVIDIA	80 GB	2,040 GB/s	Nov 2020	14.0 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: Mistral AI
Organisation type: Industry
Country: France
Published: 17 September 2024
Authors: Mistral AI Team

What it does

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

Domain: Vision, Language, Multimodal
Task: Language modeling/generation, Question answering, Visual question answering, Code generation

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: 12.4B
Training data: 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.

Likely above 10²³ FLOP: Yes
Record confidence: Confident

Sources

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

Reference: Pixtral 12B - the first-ever multimodal Mistral model. Apache 2.0.
Last updated: 28 November 2025

The extremes

The ten fastest GPUs that run Pixtral 12B

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 Pixtral 12B

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

Xeon Phi 5110P

Memory needed

6.8 GB

Fastest

273 tok/s

Pixtral 12B is small enough at 12.4B parameters that hardware is rarely the obstacle — 509 of the cards we track can run it, including cards several years old.

The smallest card that holds it is the Xeon Phi 5110P with 8 GB, running it at Q3_K_M and producing around 19.2 tokens per second.

Top of the range is the B200, at roughly 273 tokens per second thanks to 8,000 GB/s of bandwidth.

Background

Pixtral 12B was published by Mistral AI, in France, in September 2024. industry is the category the publisher falls under.

It works in Vision, Language, Multimodal, and is recorded as doing language modeling/generation, Question answering, Visual question answering, Code generation.

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

Reading the throughput figures

The median result is around 20.6 tokens per second; 453 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.

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.

Step by step

How to choose a GPU for Pixtral 12B

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

The table lists every card that can hold Pixtral 12B — around 6.8 GB at Q3_K_M. That figure, not the card's headline performance, is what decides whether it runs.
02

Set the context length you will work at

Longer conversations cost memory on top of what the weights need. Move the slider to your real working length before trusting any row for Pixtral 12B.
03

Choose how far you will compress it

The quantisation column varies by card, because a bigger card holds a more accurate copy of Pixtral 12B — Q3_K_M on the smallest card that fits. Set a floor to hold the comparison at one level.
04

Rank by throughput rather than spec sheet

Sort by speed to see how cards rank for Pixtral 12B. It will not match a gaming ordering — generation is bound by memory bandwidth, which is why the B200 tops it at 273 tok/s.
05

Read the fit column last

A tight fit runs Pixtral 12B 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

Following a card through to its own page shows every other model it can hold, which is the question that follows once Pixtral 12B is settled.

Answers

Pixtral 12B — common questions

How fast is Pixtral 12B on a GPU?

It depends on the card. The quickest we calculate is a B200 at about 273 tokens per second; the slowest that still runs it manages considerably less. Reading speed is around ten tokens per second, and 453 of the cards that can run Pixtral 12B clear that.

How much VRAM does Pixtral 12B need?

About 6.8 GB at Q3_K_M 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 Pixtral 12B on a 8 GB GPU?

Yes. A CMP 170HX 8 GB with 8 GB runs it at Q3_K_M, using about 6.8 GB and generating roughly 137 tokens per second — a tight fit.

Can I run Pixtral 12B on a 12 GB GPU?

Yes. A GeForce RTX 3080 Ti with 12 GB runs it at Q5_K_M, using about 9.6 GB and generating roughly 55.7 tokens per second — a tight fit.

Can I run Pixtral 12B on a 16 GB GPU?

Yes. A Tesla V100 SXM2 16 GB with 16 GB runs it at Q8_0, using about 14.0 GB and generating roughly 38.6 tokens per second — a tight fit.

Can I run Pixtral 12B on a 24 GB GPU?

Yes. A GeForce RTX 5090 D V2 with 24 GB runs it at Q8_0, using about 14.0 GB and generating roughly 45.8 tokens per second — a comfortable fit.

Is Pixtral 12B open source?

Its weights are published, so Pixtral 12B 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 Pixtral 12B have?

Pixtral 12B has 12.4B parameters. "New 400M parameter vision encoder trained from scratch" + "12B parameter multimodal decoder based on Mistral Nemo". 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 Pixtral 12B?

Pixtral 12B was published by Mistral AI, based in France, categorised as industry.

When was Pixtral 12B released?

Pixtral 12B was published in September 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 Pixtral 12B used for?

Pixtral 12B works in Vision, Language, Multimodal, and is recorded as handling language modeling/generation, Question answering, Visual question answering, Code generation. Models frequently carry more than one of each, and the tags describe purpose rather than capability limits.

Where can I download Pixtral 12B?

The weights for Pixtral 12B are published, though we do not hold a repository link for it. This site calculates hardware requirements rather than hosting model files.

Can I run Pixtral 12B 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 — the nearest miss we calculate is short by 2.8 GB. Our figures for Pixtral 12B assume it is fully resident.

Would two GPUs run Pixtral 12B faster?

Capacity adds across cards; throughput does not. Since 509 of the cards we track already hold Pixtral 12B on their own, a second card is rarely the answer here.

Why does the quantisation differ between cards for Pixtral 12B?

Because capacity varies, so does how hard Pixtral 12B has to be squeezed — 4 distinct levels appear in the table above. Set a minimum quality to compare at one.

How accurate are these Pixtral 12B 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 164–437 tok/s on the B200 rather than a single number.

What GPU do I need to run Pixtral 12B?

The smallest card in our catalogue that holds Pixtral 12B is the Xeon Phi 5110P, with 8 GB of memory. It runs the model at Q3_K_M using about 6.8 GB, and produces roughly 19.2 tokens per second. 509 cards in total can run it.

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