mPLUG-Owl2 TPS calculator

Open weights Alibaba 7.1B parameters November 2023

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

589 cards that can run it

818 cards we hold specifications for

Smallest card that fits

Tesla K20c

5 GB · Q3_K_M · 28.4 tok/s

Fastest card

B200

476 tok/s · 180 GB

Which GPUs can run mPLUG-Owl2?

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

589 cards match

Calculating

					Needs	Quantisation	Fit
476 tok/s 286–761 · low confidence	B200 NVIDIA	180 GB	8,000 GB/s	Jan 2024	8.3 GB	Q8_0	Comfortable
476 tok/s 286–761 · low confidence	B300 NVIDIA	288 GB	8,000 GB/s	Sep 2025	8.3 GB	Q8_0	Comfortable
380 tok/s 228–608 · low confidence	Radeon Instinct MI350X AMD	288 GB	8,190 GB/s	Jan 2025	8.3 GB	Q8_0	Comfortable
380 tok/s 228–608 · low confidence	Radeon Instinct MI355X AMD	288 GB	8,190 GB/s	Jan 2025	8.3 GB	Q8_0	Comfortable
304 tok/s 182–486 · low confidence	Radeon Instinct MI300 AMD	128 GB	6,550 GB/s	Jan 2023	8.3 GB	Q8_0	Comfortable
291 tok/s 175–465 · low confidence	H200 NVL NVIDIA	141 GB	4,890 GB/s	Nov 2024	8.3 GB	Q8_0	Comfortable
291 tok/s 175–465 · low confidence	H200 SXM 141 GB NVIDIA	141 GB	4,890 GB/s	Nov 2024	8.3 GB	Q8_0	Comfortable
278 tok/s 167–445 · low confidence	Radeon Instinct MI325X AMD	256 GB	6,000 GB/s	Oct 2024	8.3 GB	Q8_0	Comfortable
247 tok/s 148–395 · low confidence	Radeon Instinct MI300A AMD	128 GB	5,325 GB/s	Dec 2023	8.3 GB	Q8_0	Comfortable
247 tok/s 148–395 · low confidence	Radeon Instinct MI300X AMD	192 GB	5,325 GB/s	Dec 2023	8.3 GB	Q8_0	Comfortable
247 tok/s 148–395 · low confidence	Radeon Instinct MI308X AMD	192 GB	5,325 GB/s	Dec 2023	8.3 GB	Q8_0	Comfortable
234 tok/s 141–375 · low confidence	H100 NVL 94 GB NVIDIA	94 GB	3,940 GB/s	Mar 2023	8.3 GB	Q8_0	Comfortable
200 tok/s 120–320 · low confidence	H100 PCIe 96 GB NVIDIA	96 GB	3,360 GB/s	Mar 2023	8.3 GB	Q8_0	Comfortable
200 tok/s 120–320 · low confidence	H100 SXM5 80 GB NVIDIA	80 GB	3,360 GB/s	Oct 2022	8.3 GB	Q8_0	Comfortable
200 tok/s 120–320 · low confidence	H100 SXM5 94 GB NVIDIA	94 GB	3,360 GB/s	Mar 2023	8.3 GB	Q8_0	Comfortable
200 tok/s 120–320 · low confidence	H100 SXM5 96 GB NVIDIA	96 GB	3,360 GB/s	Mar 2023	8.3 GB	Q8_0	Comfortable
200 tok/s 120–320 · low confidence	H800 SXM5 NVIDIA	80 GB	3,360 GB/s	Mar 2023	8.3 GB	Q8_0	Comfortable
152 tok/s 91–244 · low confidence	Radeon Instinct MI250 AMD	128 GB	3,280 GB/s	Nov 2021	8.3 GB	Q8_0	Comfortable
152 tok/s 91–244 · low confidence	Radeon Instinct MI250X AMD	128 GB	3,280 GB/s	Nov 2021	8.3 GB	Q8_0	Comfortable
129 tok/s 77–206 · low confidence	CMP 170HX 8 GB NVIDIA	8 GB	1,490 GB/s	Sep 2021	6.7 GB	Q6_K	Tight
127 tok/s 76–203 · low confidence	Data Center GPU Max 1550 Intel	128 GB	3,280 GB/s	Jan 2023	8.3 GB	Q8_0	Comfortable
124 tok/s 74–199 · low confidence	Data Center GPU Max Subsystem Intel	128 GB	3,210 GB/s	Jan 2023	8.3 GB	Q8_0	Comfortable
121 tok/s 73–194 · low confidence	A100 SXM4 80 GB NVIDIA	80 GB	2,040 GB/s	Nov 2020	8.3 GB	Q8_0	Comfortable
121 tok/s 73–194 · low confidence	A100X NVIDIA	80 GB	2,040 GB/s	Jun 2021	8.3 GB	Q8_0	Comfortable
121 tok/s 73–194 · low confidence	A800 SXM4 80 GB NVIDIA	80 GB	2,040 GB/s	Aug 2022	8.3 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: Alibaba
Organisation type: Industry
Country: China
Published: 7 November 2023
Authors: Qinghao Ye, Haiyang Xu, Jiabo Ye, Ming Yan, Anwen Hu, Haowei Liu, Qi Qian, Ji Zhang, Fei Huang, Jingren Zhou

What it does

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

Domain: Vision, Language, Multimodal
Task: Visual question answering, Image captioning, Language modeling/generation
Base model: Llama 2-7B
Numerical format: BF16

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: 7.1B
Training data: 182,032,793,600 tokens
Epochs: 1

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.

How it was established: Operation counting
Fine-tuning compute: 1.7 × 10¹⁹ 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)
Training code: Open source
Hugging Face: Mizukiluke

How it is classified

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

Why it is tracked: SOTA improvement
Record confidence: Speculative

Sources

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

Reference: mPLUG-Owl2: Revolutionizing Multi-modal Large Language Model with Modality Collaboration
Last updated: 28 November 2025

The extremes

The ten fastest GPUs that run mPLUG-Owl2

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 mPLUG-Owl2

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

Tesla K20c

Memory needed

4.2 GB

Fastest

476 tok/s

mPLUG-Owl2 reaches a parameter count of 7.1B. That is small enough that hardware is rarely the obstacle, including on cards several years old. The number of cards we track that can run it: 589.

At the low end it is handled by Tesla K20c, with a memory capacity of 5 GB, running it at a compression of Q3_K_M and producing around 28.4 tokens per second.

Top of the range is B200, generating roughly 476 tokens per second on the strength of a memory bandwidth of 8,000 GB/s.

Background

mPLUG-Owl2 was published by Alibaba, in the country recorded as China, during November 2023. The publishing organisation is categorised as industry.

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

Its starting point was an existing base model, Llama 2-7B. That is the usual way a specialised model is produced.

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 Mizukiluke.

Reading the throughput figures

Half the cards that hold it manage more than 25.7 tokens per second. Producing text faster than most people read it: 559 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.

What went into building it

It was trained on a corpus of about 182,032,793,600 tokens of text.

The reason it appears in this catalogue at all: sOTA improvement.

Step by step

How to choose a GPU for mPLUG-Owl2

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

01

Check what it needs before anything else

Start from what it actually needs, which is the requirement of mPLUG-Owl2, needing around 4.2 GB at a compression of Q3_K_M. Capacity is the gate — a card either holds it or it does not.
02

Set the context length you will work at

The conversation occupies memory too, and grows as it goes. Set the slider to the length you expect, because at long context a card that handles short questions easily can be dropped by mPLUG-Owl2.
03

Decide how much compression you will accept

The quantisation column varies by card, because a bigger card holds a more accurate copy, 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

Compare tokens per second, not specifications

The speed ordering is effectively an ordering by memory bandwidth, for mPLUG-Owl2. It will not match a gaming ordering, because generation is bound by memory bandwidth. The card topping the list is B200, at 476 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 mPLUG-Owl2. 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

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 mPLUG-Owl2.

Answers

mPLUG-Owl2 — common questions

mPLUG-Owl2— would two GPUs run it faster?

Capacity adds across cards; throughput does not. The number of cards already holding it on their own: 589. So a second card is rarely the answer here.

mPLUG-Owl2— why does the quantisation differ between cards?

A larger card holds a more accurate copy. The number of compression levels used across the cards that run it: 4. Bigger cards get the more accurate version, and the quality floor above pins the comparison to one level.

mPLUG-Owl2— how accurate are these speed estimates?

They are calculated from specifications rather than measured, and each carries a range. One example: 286–761 tok/s on B200. The same model and card vary by thirty to fifty per cent depending on the inference software and its version.

mPLUG-Owl2— what GPU do I need to run it?

The smallest card in our catalogue that holds it is Tesla K20c, with a memory capacity of 5 GB. It runs the model at a compression of Q3_K_M using about 4.2 GB, and produces roughly 28.4 tokens per second. The number of cards able to run it in total: 589.

mPLUG-Owl2— how fast is it on a GPU?

It depends on the card. The quickest we calculate is B200, at about 476 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: 559.

mPLUG-Owl2— how much VRAM does it need?

It needs about 4.2 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.

mPLUG-Owl2— can I run it on a GPU holding 8 GB?

Yes. The card CMP 170HX 8 GB, holding 8 GB, runs it at a compression of Q6_K, using about 6.7 GB and generating roughly 129 tokens per second. The fit is tight.

mPLUG-Owl2— can I run it on a GPU holding 12 GB?

Yes. The card GeForce RTX 3080 Ti, holding 12 GB, runs it at a compression of Q8_0, using about 8.3 GB and generating roughly 54.3 tokens per second. The fit is comfortable.

mPLUG-Owl2— 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 Q8_0, using about 8.3 GB and generating roughly 67.2 tokens per second. The fit is comfortable.

mPLUG-Owl2— 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 Q8_0, using about 8.3 GB and generating roughly 79.7 tokens per second. The fit is comfortable.

mPLUG-Owl2— 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.

mPLUG-Owl2— how many parameters does it have?

It has a parameter count of 7.1B. "As depicted in Figure 2, our model, referred to as mPLUGOwl2, is composed of three main components: a fundamental vision encoder, a visual abstractor, and a language decoder. Specifically, we utilize ViT-L/14 as the vision encoder and LLaMA-2-7B [58] as the language decoder" ViT-L/14 has 123M parameters and Llama 2 7B has 7B parameters. 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.

mPLUG-Owl2— who created it?

It was published by Alibaba, based in China, an organisation categorised as industry.

mPLUG-Owl2— when was it released?

It was published in November 2023. Capability per parameter has improved considerably since, so a newer model of the same size is often the better use of the same hardware.

mPLUG-Owl2— what is it used for?

It works in the domain of Vision, Language, Multimodal, and is recorded as handling the task of visual question answering, Image captioning, Language modeling/generation. A model can carry several of each, so these are the areas it was built for rather than a limit on what it will attempt.

mPLUG-Owl2— where can I download it?

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

mPLUG-Owl2— can I run it if it does not fit in my GPU?

Partly. Layers that do not fit sit in system memory and run at a fraction of the speed, so a mostly-offloaded model is rarely worth using. The nearest miss we calculate falls short by 1.4 GB. Every figure here assumes the whole model is resident on the card.

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