Aya Vision 32B TPS calculator

Open weights Cohere 33.1B parameters March 2025

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 · Q3_K_M · 22.1 tok/s

Fastest card

B200

102 tok/s · 180 GB

Which GPUs can run Aya Vision 32B?

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
102 tok/s 61–164 · low confidence	B200 NVIDIA	180 GB	8,000 GB/s	Jan 2024	36.1 GB	Q8_0	Comfortable
102 tok/s 61–164 · low confidence	B300 NVIDIA	288 GB	8,000 GB/s	Sep 2025	36.1 GB	Q8_0	Comfortable
81.7 tok/s 49–131 · low confidence	Radeon Instinct MI350X AMD	288 GB	8,190 GB/s	Jan 2025	36.1 GB	Q8_0	Comfortable
81.7 tok/s 49–131 · low confidence	Radeon Instinct MI355X AMD	288 GB	8,190 GB/s	Jan 2025	36.1 GB	Q8_0	Comfortable
65.4 tok/s 39–105 · low confidence	Radeon Instinct MI300 AMD	128 GB	6,550 GB/s	Jan 2023	36.1 GB	Q8_0	Comfortable
62.6 tok/s 38–100 · low confidence	H200 NVL NVIDIA	141 GB	4,890 GB/s	Nov 2024	36.1 GB	Q8_0	Comfortable
62.6 tok/s 38–100 · low confidence	H200 SXM 141 GB NVIDIA	141 GB	4,890 GB/s	Nov 2024	36.1 GB	Q8_0	Comfortable
59.9 tok/s 36–96 · low confidence	Radeon Instinct MI325X AMD	256 GB	6,000 GB/s	Oct 2024	36.1 GB	Q8_0	Comfortable
53.2 tok/s 32–85 · low confidence	Radeon Instinct MI300A AMD	128 GB	5,325 GB/s	Dec 2023	36.1 GB	Q8_0	Comfortable
53.2 tok/s 32–85 · low confidence	Radeon Instinct MI300X AMD	192 GB	5,325 GB/s	Dec 2023	36.1 GB	Q8_0	Comfortable
53.2 tok/s 32–85 · low confidence	Radeon Instinct MI308X AMD	192 GB	5,325 GB/s	Dec 2023	36.1 GB	Q8_0	Comfortable
50.4 tok/s 30–81 · low confidence	H100 NVL 94 GB NVIDIA	94 GB	3,940 GB/s	Mar 2023	36.1 GB	Q8_0	Comfortable
43.0 tok/s 26–69 · low confidence	H100 PCIe 96 GB NVIDIA	96 GB	3,360 GB/s	Mar 2023	36.1 GB	Q8_0	Comfortable
43.0 tok/s 26–69 · low confidence	H100 SXM5 80 GB NVIDIA	80 GB	3,360 GB/s	Oct 2022	36.1 GB	Q8_0	Comfortable
43.0 tok/s 26–69 · low confidence	H100 SXM5 94 GB NVIDIA	94 GB	3,360 GB/s	Mar 2023	36.1 GB	Q8_0	Comfortable
43.0 tok/s 26–69 · low confidence	H100 SXM5 96 GB NVIDIA	96 GB	3,360 GB/s	Mar 2023	36.1 GB	Q8_0	Comfortable
43.0 tok/s 26–69 · low confidence	H800 SXM5 NVIDIA	80 GB	3,360 GB/s	Mar 2023	36.1 GB	Q8_0	Comfortable
39.6 tok/s 24–63 · low confidence	GeForce RTX 5090 D V2 NVIDIA	24 GB	1,340 GB/s	Aug 2025	20.7 GB	Q4_K_M	Tight
36.0 tok/s 22–58 · low confidence	A30X NVIDIA	24 GB	1,220 GB/s	Apr 2021	20.7 GB	Q4_K_M	Tight
34.8 tok/s 21–56 · low confidence	DRIVE A100 PROD NVIDIA	32 GB	1,870 GB/s	May 2020	28.4 GB	Q6_K	Tight
34.8 tok/s 21–56 · low confidence	GRID A100A NVIDIA	32 GB	1,870 GB/s	May 2020	28.4 GB	Q6_K	Tight
33.3 tok/s 20–53 · low confidence	GeForce RTX 5090 NVIDIA	32 GB	1,790 GB/s	Jan 2025	28.4 GB	Q6_K	Tight
33.3 tok/s 20–53 · low confidence	GeForce RTX 5090 D NVIDIA	32 GB	1,790 GB/s	Jan 2025	28.4 GB	Q6_K	Tight
32.7 tok/s 20–52 · low confidence	Radeon Instinct MI250 AMD	128 GB	3,280 GB/s	Nov 2021	36.1 GB	Q8_0	Comfortable
32.7 tok/s 20–52 · low confidence	Radeon Instinct MI250X AMD	128 GB	3,280 GB/s	Nov 2021	36.1 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: Cohere
Organisation type: Industry
Country: Canada
Published: 3 March 2025
Authors: Cohere Labs Team

What it does

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

Domain: Multimodal, Vision, Language
Task: Language generation, Image captioning, Visual question answering, Language modeling/generation, Question answering
Base model: Aya Expanse 32B

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: 33.1B
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 (non-commercial)
Training code: Unreleased
Hugging Face: CohereLabs

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.

Reference: Aya Vision: Expanding the worlds AI can see
Last updated: 28 November 2025

The extremes

The ten fastest GPUs that run Aya Vision 32B

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 Aya Vision 32B

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

The hardware side

Minimum card

RTX A4500

Memory needed

16.9 GB

Fastest

102 tok/s

Aya Vision 32B reaches a parameter count of 33.1B. 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.

At the low end it is handled by RTX A4500, with a memory capacity of 20 GB, running it at a compression of Q3_K_M and producing around 22.1 tokens per second.

At the other end sits B200, generating roughly 102 tokens per second on the strength of a memory bandwidth of 8,000 GB/s.

Where it came from

Aya Vision 32B was published by Cohere, in the country recorded as Canada, during March 2025. The publishing organisation is categorised as industry.

It works in the domain of Multimodal, Vision, Language, and is recorded as performing the task of language generation, Image captioning, Visual question answering, Language modeling/generation, Question answering.

Its starting point was an existing base model, Aya Expanse 32B. That is why it shares the base model's general shape and size.

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

Understanding the speeds

Across every card that can run it, the middle of the range sits at 20.3 tokens per second. Clearing the ten tokens per second that roughly matches reading speed: 101 of them.

Being dense, it reads all of itself per token, which is why the ordering by speed below follows the ordering by memory bandwidth so closely.

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 Aya Vision 32B

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

The table lists every card able to hold Aya Vision 32B, needing around 16.9 GB at a compression of Q3_K_M. Capacity is the gate — a card either holds it or it does not.
02

Decide how long your conversations run

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 Aya Vision 32B.
03

Choose how far you will compress it

Compression is what makes a model fit smaller cards, at some cost in accuracy, reaching a compression of Q3_K_M 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

Sort by speed

Sort by speed to see how cards rank for Aya Vision 32B. It will not match a gaming ordering, because generation is bound by memory bandwidth. The card topping the list is B200, at 102 tok/s.
05

Check the fit verdict before buying

Tight means it loads and works with no room to raise the context later, in the case of Aya Vision 32B. 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

Every card name links to its own page, which runs the same calculation across the whole model catalogue. A card is usually bought for more than one model, so it is worth a look before buying for Aya Vision 32B.

Answers

Aya Vision 32B — common questions

Aya Vision 32B— how many parameters does it have?

It has a parameter count of 33.1B. According to huggingface.co/CohereLabs/aya-vision-32b. 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.

Aya Vision 32B— who created it?

It was published by Cohere, based in Canada, an organisation categorised as industry.

Aya Vision 32B— when was it released?

It was published in March 2025.

Aya Vision 32B— what is it used for?

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

Aya Vision 32B— where can I download it?

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

Aya Vision 32B— 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 6.3 GB. Every figure here assumes the whole model is resident on the card.

Aya Vision 32B— 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.

Aya Vision 32B— 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.

Aya Vision 32B— 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: 61–164 tok/s on B200. The same model and card vary by thirty to fifty per cent depending on the inference software and its version.

Aya Vision 32B— 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 Q3_K_M using about 16.9 GB, and produces roughly 22.1 tokens per second. The number of cards able to run it in total: 132.

Aya Vision 32B— how fast is it on a GPU?

It depends on the card. The quickest we calculate is B200, at about 102 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: 101.

Aya Vision 32B— how much VRAM does it need?

It needs about 16.9 GB at a compression of Q3_K_M, 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.

Aya Vision 32B— 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 20.7 GB and generating roughly 39.6 tokens per second. The fit is tight.

Aya Vision 32B— 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.

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