Gemma 4 26B A4B TPS calculator

Open weights Google DeepMind 25.2B 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

241 cards that can run it

818 cards we hold specifications for

Smallest card that fits

Xeon Phi 7120P

16 GB · IQ4_XS · 52.5 tok/s

Fastest card

B200

747 tok/s · 180 GB

Which GPUs can run Gemma 4 26B A4B?

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

241 cards match

Calculating

					Needs	Quantisation	Fit
747 tok/s 448–1,195 · low confidence	B200 NVIDIA	180 GB	8,000 GB/s	Jan 2024	26.7 GB	Q8_0	Comfortable
747 tok/s 448–1,195 · low confidence	B300 NVIDIA	288 GB	8,000 GB/s	Sep 2025	26.7 GB	Q8_0	Comfortable
596 tok/s 358–954 · low confidence	Radeon Instinct MI350X AMD	288 GB	8,190 GB/s	Jan 2025	26.7 GB	Q8_0	Comfortable
596 tok/s 358–954 · low confidence	Radeon Instinct MI355X AMD	288 GB	8,190 GB/s	Jan 2025	26.7 GB	Q8_0	Comfortable
477 tok/s 286–763 · low confidence	Radeon Instinct MI300 AMD	128 GB	6,550 GB/s	Jan 2023	26.7 GB	Q8_0	Comfortable
457 tok/s 274–731 · low confidence	H200 NVL NVIDIA	141 GB	4,890 GB/s	Nov 2024	26.7 GB	Q8_0	Comfortable
457 tok/s 274–731 · low confidence	H200 SXM 141 GB NVIDIA	141 GB	4,890 GB/s	Nov 2024	26.7 GB	Q8_0	Comfortable
437 tok/s 262–699 · low confidence	Radeon Instinct MI325X AMD	256 GB	6,000 GB/s	Oct 2024	26.7 GB	Q8_0	Comfortable
388 tok/s 233–621 · low confidence	Radeon Instinct MI300A AMD	128 GB	5,325 GB/s	Dec 2023	26.7 GB	Q8_0	Comfortable
388 tok/s 233–621 · low confidence	Radeon Instinct MI300X AMD	192 GB	5,325 GB/s	Dec 2023	26.7 GB	Q8_0	Comfortable
388 tok/s 233–621 · low confidence	Radeon Instinct MI308X AMD	192 GB	5,325 GB/s	Dec 2023	26.7 GB	Q8_0	Comfortable
368 tok/s 221–589 · low confidence	H100 NVL 94 GB NVIDIA	94 GB	3,940 GB/s	Mar 2023	26.7 GB	Q8_0	Comfortable
314 tok/s 188–502 · low confidence	H100 PCIe 96 GB NVIDIA	96 GB	3,360 GB/s	Mar 2023	26.7 GB	Q8_0	Comfortable
314 tok/s 188–502 · low confidence	H100 SXM5 80 GB NVIDIA	80 GB	3,360 GB/s	Oct 2022	26.7 GB	Q8_0	Comfortable
314 tok/s 188–502 · low confidence	H100 SXM5 94 GB NVIDIA	94 GB	3,360 GB/s	Mar 2023	26.7 GB	Q8_0	Comfortable
314 tok/s 188–502 · low confidence	H100 SXM5 96 GB NVIDIA	96 GB	3,360 GB/s	Mar 2023	26.7 GB	Q8_0	Comfortable
314 tok/s 188–502 · low confidence	H800 SXM5 NVIDIA	80 GB	3,360 GB/s	Mar 2023	26.7 GB	Q8_0	Comfortable
259 tok/s 155–415 · low confidence	Tesla V100 SXM2 16 GB NVIDIA	16 GB	1,130 GB/s	Nov 2019	13.5 GB	IQ4_XS	Tight
239 tok/s 143–382 · low confidence	Radeon Instinct MI250 AMD	128 GB	3,280 GB/s	Nov 2021	26.7 GB	Q8_0	Comfortable
239 tok/s 143–382 · low confidence	Radeon Instinct MI250X AMD	128 GB	3,280 GB/s	Nov 2021	26.7 GB	Q8_0	Comfortable
220 tok/s 132–352 · low confidence	GeForce RTX 5080 NVIDIA	16 GB	960 GB/s	Jan 2025	13.5 GB	IQ4_XS	Tight
206 tok/s 123–329 · low confidence	Tesla V100 DGXS 16 GB NVIDIA	16 GB	897 GB/s	Mar 2018	13.5 GB	IQ4_XS	Tight
206 tok/s 123–329 · low confidence	Tesla V100 PCIe 16 GB NVIDIA	16 GB	897 GB/s	Jun 2017	13.5 GB	IQ4_XS	Tight
206 tok/s 123–329 · low confidence	Tesla V100 SXM2 16 GB NVIDIA	16 GB	897 GB/s	Jun 2017	13.5 GB	IQ4_XS	Tight
205 tok/s 123–329 · low confidence	GeForce RTX 5070 Ti NVIDIA	16 GB	896 GB/s	Feb 2025	13.5 GB	IQ4_XS	Tight

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.

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: 25.2B
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
Hugging Face: google

How it is classified

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

Record confidence: Confident

Sources

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

Last updated: 19 June 2026

The extremes

The ten fastest GPUs that run Gemma 4 26B A4B

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 26B A4B

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 7120P

Memory needed

13.5 GB

Fastest

747 tok/s

Gemma 4 26B A4B reaches a parameter count of 25.2B. 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: 241.

The smallest card that holds it is Xeon Phi 7120P, with a memory capacity of 16 GB, running it at a compression of IQ4_XS and producing around 52.5 tokens per second.

At the other end sits B200, generating roughly 747 tokens per second on the strength of a memory bandwidth of 8,000 GB/s.

Background

Gemma 4 26B A4B was published by Google DeepMind, in the country recorded as United States of America, during April 2026. The publishing organisation is categorised as industry.

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. On Hugging Face it is published under the organisation google.

Reading the throughput figures

Across every card that can run it, the middle of the range sits at 102.7 tokens per second. Producing text faster than most people read it: 239 of them.

This is a mixture-of-experts model, which routes each token through only part of itself. It therefore generates far faster than its total size suggests — while still needing every parameter resident in memory, so it is quick without being cheap to hold.

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.

Step by step

How to choose a GPU for Gemma 4 26B A4B

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

Start from what it actually needs, which is the requirement of Gemma 4 26B A4B, needing around 13.5 GB at a compression of IQ4_XS. 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 a card that seemed fine stops fitting Gemma 4 26B A4B.
03

Set a quality floor

Each card runs the least-compressed copy it can hold, 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

Compare tokens per second, not specifications

The speed ordering is effectively an ordering by memory bandwidth, for Gemma 4 26B A4B. It will not match a gaming ordering, because generation is bound by memory bandwidth. The card topping the list is B200, at 747 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 26B A4B. 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

See what else that card runs

Each card page repeats this sweep for every model we hold, answering what else the hardware is good for beyond Gemma 4 26B A4B.

Answers

Gemma 4 26B A4B — common questions

Gemma 4 26B A4B— can I run it on a GPU holding 16 GB?

Yes. The card Tesla V100 SXM2 16 GB, holding 16 GB, runs it at a compression of IQ4_XS, using about 13.5 GB and generating roughly 259 tokens per second. The fit is tight.

Gemma 4 26B A4B— 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 Q6_K, using about 20.8 GB and generating roughly 182 tokens per second. The fit is tight.

Gemma 4 26B A4B— 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 26B A4B— how many parameters does it have?

It has a parameter count of 25.2B. 25.2B total, 3.8B active. 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 26B A4B— who created it?

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

Gemma 4 26B A4B— when was it released?

It was published in April 2026.

Gemma 4 26B A4B— what is it used for?

and is recorded as handling the task of language modeling/generation, Question answering. Models frequently carry more than one of each, and the tags describe purpose rather than capability limits.

Gemma 4 26B A4B— where can I download it?

Its weights are published on Hugging Face, under the organisation google. We do not host model files — this site calculates what hardware is needed to run them.

Gemma 4 26B A4B— 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 4.2 GB. Every figure here assumes the whole model is resident on the card.

Gemma 4 26B A4B— would two GPUs run it faster?

Capacity adds across cards; throughput does not. The number of cards already holding it on their own: 241. So a second card is rarely the answer here.

Gemma 4 26B A4B— 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: 4. Bigger cards get the more accurate version, and the quality floor above pins the comparison to one level.

Gemma 4 26B A4B— how accurate are these speed estimates?

These are estimates with real error bars, and any of them could reasonably land anywhere in its published range depending on which runtime you use. The fastest result here: 448–1,195 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 26B A4B— what GPU do I need to run it?

The smallest card in our catalogue that holds it is Xeon Phi 7120P, with a memory capacity of 16 GB. It runs the model at a compression of IQ4_XS using about 13.5 GB, and produces roughly 52.5 tokens per second. The number of cards able to run it in total: 241.

Gemma 4 26B A4B— how fast is it on a GPU?

It depends on the card. The quickest we calculate is B200, at about 747 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: 239.

Gemma 4 26B A4B— how much VRAM does it need?

It needs about 13.5 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.

Source

Original publication

Record last updated 19 June 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