Gemma 4 31B IT TPS calculator

Open weights Google DeepMind 31B parameters April 2026

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

132 cards that can run it

818 cards we hold specifications for

Smallest card that fits

RTX A4500

20 GB · IQ4_XS · 21.5 tok/s

Fastest card

B200

109 tok/s · 180 GB

Which GPUs can run Gemma 4 31B IT?

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

132 cards match

Calculating

					Needs	Quantisation	Fit
109 tok/s 66–175 · low confidence	B200 NVIDIA	180 GB	8,000 GB/s	Jan 2024	33.9 GB	Q8_0	Comfortable
109 tok/s 66–175 · low confidence	B300 NVIDIA	288 GB	8,000 GB/s	Sep 2025	33.9 GB	Q8_0	Comfortable
87.3 tok/s 52–140 · low confidence	Radeon Instinct MI350X AMD	288 GB	8,190 GB/s	Jan 2025	33.9 GB	Q8_0	Comfortable
87.3 tok/s 52–140 · low confidence	Radeon Instinct MI355X AMD	288 GB	8,190 GB/s	Jan 2025	33.9 GB	Q8_0	Comfortable
69.8 tok/s 42–112 · low confidence	Radeon Instinct MI300 AMD	128 GB	6,550 GB/s	Jan 2023	33.9 GB	Q8_0	Comfortable
66.8 tok/s 40–107 · low confidence	H200 NVL NVIDIA	141 GB	4,890 GB/s	Nov 2024	33.9 GB	Q8_0	Comfortable
66.8 tok/s 40–107 · low confidence	H200 SXM 141 GB NVIDIA	141 GB	4,890 GB/s	Nov 2024	33.9 GB	Q8_0	Comfortable
63.9 tok/s 38–102 · low confidence	Radeon Instinct MI325X AMD	256 GB	6,000 GB/s	Oct 2024	33.9 GB	Q8_0	Comfortable
56.8 tok/s 34–91 · low confidence	Radeon Instinct MI300A AMD	128 GB	5,325 GB/s	Dec 2023	33.9 GB	Q8_0	Comfortable
56.8 tok/s 34–91 · low confidence	Radeon Instinct MI300X AMD	192 GB	5,325 GB/s	Dec 2023	33.9 GB	Q8_0	Comfortable
56.8 tok/s 34–91 · low confidence	Radeon Instinct MI308X AMD	192 GB	5,325 GB/s	Dec 2023	33.9 GB	Q8_0	Comfortable
53.8 tok/s 32–86 · low confidence	H100 NVL 94 GB NVIDIA	94 GB	3,940 GB/s	Mar 2023	33.9 GB	Q8_0	Comfortable
45.9 tok/s 28–73 · low confidence	H100 PCIe 96 GB NVIDIA	96 GB	3,360 GB/s	Mar 2023	33.9 GB	Q8_0	Comfortable
45.9 tok/s 28–73 · low confidence	H100 SXM5 80 GB NVIDIA	80 GB	3,360 GB/s	Oct 2022	33.9 GB	Q8_0	Comfortable
45.9 tok/s 28–73 · low confidence	H100 SXM5 94 GB NVIDIA	94 GB	3,360 GB/s	Mar 2023	33.9 GB	Q8_0	Comfortable
45.9 tok/s 28–73 · low confidence	H100 SXM5 96 GB NVIDIA	96 GB	3,360 GB/s	Mar 2023	33.9 GB	Q8_0	Comfortable
45.9 tok/s 28–73 · low confidence	H800 SXM5 NVIDIA	80 GB	3,360 GB/s	Mar 2023	33.9 GB	Q8_0	Comfortable
42.3 tok/s 25–68 · low confidence	GeForce RTX 5090 D V2 NVIDIA	24 GB	1,340 GB/s	Aug 2025	19.5 GB	Q4_K_M	Tight
38.5 tok/s 23–62 · low confidence	A30X NVIDIA	24 GB	1,220 GB/s	Apr 2021	19.5 GB	Q4_K_M	Tight
37.1 tok/s 22–59 · low confidence	DRIVE A100 PROD NVIDIA	32 GB	1,870 GB/s	May 2020	26.7 GB	Q6_K	Tight
37.1 tok/s 22–59 · low confidence	GRID A100A NVIDIA	32 GB	1,870 GB/s	May 2020	26.7 GB	Q6_K	Tight
35.5 tok/s 21–57 · low confidence	GeForce RTX 5090 NVIDIA	32 GB	1,790 GB/s	Jan 2025	26.7 GB	Q6_K	Tight
35.5 tok/s 21–57 · low confidence	GeForce RTX 5090 D NVIDIA	32 GB	1,790 GB/s	Jan 2025	26.7 GB	Q6_K	Tight
35.0 tok/s 21–56 · low confidence	Radeon Instinct MI250 AMD	128 GB	3,280 GB/s	Nov 2021	33.9 GB	Q8_0	Comfortable
35.0 tok/s 21–56 · low confidence	Radeon Instinct MI250X AMD	128 GB	3,280 GB/s	Nov 2021	33.9 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: Google DeepMind
Organisation type: Industry
Country: United States of America
Published: 2 April 2026

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: 31B
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 (restricted use)

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.

Last updated: 8 April 2026

The extremes

The ten fastest GPUs that run Gemma 4 31B IT

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 Gemma 4 31B IT

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 you need to run it

Minimum card

RTX A4500

Memory needed

17.6 GB

Fastest

109 tok/s

Gemma 4 31B IT reaches a parameter count of 31B. That lands in the range a serious desktop card can handle once the weights are compressed. The number of cards we track that can run it: 132.

The least hardware that works is RTX A4500, with a memory capacity of 20 GB, running it at a compression of IQ4_XS and producing around 21.5 tokens per second.

Top of the range is B200, generating roughly 109 tokens per second on the strength of a memory bandwidth of 8,000 GB/s.

Where it came from

Gemma 4 31B IT was published by Google DeepMind, in the country recorded as United States of America, during April 2026. The category the publisher falls under is industry.

It works in the domain of Language, and is recorded as performing the task of language modeling/generation, Question answering.

Because its weights were released, nothing about running it depends on a provider staying available — it is yours once downloaded.

Understanding the speeds

The median result is around 21.3 tokens per second. Clearing the ten tokens per second that roughly matches reading speed: 102 of them.

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.

Its internal architecture is not on file, so memory is approximated from the parameter count and marked accordingly. Expect the real figure to differ, more so at long context.

Step by step

How to choose a GPU for Gemma 4 31B IT

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

Start from what it actually needs, which is the requirement of Gemma 4 31B IT, needing around 17.6 GB at a compression of IQ4_XS. Capacity is the gate — a card either holds it or it does not.
02

Set the context length you will work at

Set the context to what you will actually use. The cache grows with the conversation, and it is the usual reason a card that seemed fine stops fitting Gemma 4 31B IT.
03

Choose how far you will compress it

The quantisation column varies by card, because a bigger card holds a more accurate copy, reaching a compression of IQ4_XS on the smallest card that fits. Setting a minimum quality drops the cards that only manage it by squeezing further than you would want, and holds the comparison at one level.
04

Rank by throughput rather than spec sheet

The speed ordering is effectively an ordering by memory bandwidth, for Gemma 4 31B IT. It will not match a gaming ordering, because generation is bound by memory bandwidth. The card topping the list is B200, at 109 tok/s.
05

Look at the headroom, not just the fit

Tight means it loads and works with no room to raise the context later, in the case of Gemma 4 31B IT. Comfortable means you can grow the context later. That difference matters more than a few tokens per second, so buy for comfortable if you expect to.
06

Open the card you have settled on

Each card page repeats this sweep for every model we hold, answering what else the hardware is good for beyond Gemma 4 31B IT.

Answers

Gemma 4 31B IT — common questions

Gemma 4 31B IT— would two GPUs run it faster?

A second card roughly doubles the memory available but not the generation rate. The number already able to run it alone: 132. So a second card is rarely the answer here.

Gemma 4 31B IT— why does the quantisation differ between cards?

Each card is shown running the least-compressed copy it can hold. The number of distinct compression levels across the cards that fit it: 5. Bigger cards get the more accurate version, and the quality floor above pins the comparison to one level.

Gemma 4 31B IT— how accurate are these speed estimates?

They are calculated from specifications rather than measured, and each carries a range. One example: 66–175 tok/s on B200. The same model and card vary by thirty to fifty per cent depending on the inference software and its version.

Gemma 4 31B IT— what GPU do I need to run it?

The smallest card in our catalogue that holds it is RTX A4500, with a memory capacity of 20 GB. It runs the model at a compression of IQ4_XS using about 17.6 GB, and produces roughly 21.5 tokens per second. The number of cards able to run it in total: 132.

Gemma 4 31B IT— how fast is it on a GPU?

It depends on the card. The quickest we calculate is B200, at about 109 tokens per second; the slowest that still runs it manages considerably less. Reading speed is around ten tokens per second, and the number of cards clearing that: 102.

Gemma 4 31B IT— how much VRAM does it need?

It needs about 17.6 GB at a compression of IQ4_XS, 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.

Gemma 4 31B IT— can I run it on a GPU holding 24 GB?

Yes. The card GeForce RTX 5090 D V2, holding 24 GB, runs it at a compression of Q4_K_M, using about 19.5 GB and generating roughly 42.3 tokens per second. The fit is tight.

Gemma 4 31B IT— is it open source?

Its weights are published, so it 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.

Gemma 4 31B IT— how many parameters does it have?

It has a parameter count of 31B. 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.

Gemma 4 31B IT— who created it?

It was published by Google DeepMind, based in United States of America, an organisation categorised as industry.

Gemma 4 31B IT— when was it released?

It was published in April 2026.

Gemma 4 31B IT— what is it used for?

It works in the domain of Language, and is recorded as handling the task of language modeling/generation, Question answering. These are the areas it was designed around; they describe intent rather than a hard boundary.

Gemma 4 31B IT— where can I download it?

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

Gemma 4 31B IT— can I run it if it does not fit in my GPU?

It can be split between the card and system memory, but it generates painfully slowly that way. The nearest miss we calculate falls short by 5.1 GB. Every figure here assumes the whole model is resident on the card.

Source

Record last updated 8 April 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