GLM-4.5V TPS calculator

Open weights Z.ai (Zhipu AI),Tsinghua University 108B parameters August 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

43 cards that can run it

818 cards we hold specifications for

Smallest card that fits

Radeon Instinct MI200

64 GB · IQ4_XS · 68.4 tok/s

Fastest card

B200

174 tok/s · 180 GB

Which GPUs can run GLM-4.5V?

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

43 cards match

Calculating

					Needs	Quantisation	Fit
174 tok/s 105–279 · low confidence	B200 NVIDIA	180 GB	8,000 GB/s	Jan 2024	110.8 GB	Q8_0	Comfortable
174 tok/s 105–279 · low confidence	B300 NVIDIA	288 GB	8,000 GB/s	Sep 2025	110.8 GB	Q8_0	Comfortable
169 tok/s 101–270 · low confidence	H100 SXM5 80 GB NVIDIA	80 GB	3,360 GB/s	Oct 2022	60.5 GB	Q4_K_M	Tight
169 tok/s 101–270 · low confidence	H800 SXM5 NVIDIA	80 GB	3,360 GB/s	Mar 2023	60.5 GB	Q4_K_M	Tight
153 tok/s 92–245 · low confidence	H100 NVL 94 GB NVIDIA	94 GB	3,940 GB/s	Mar 2023	73.1 GB	Q5_K_M	Tight
139 tok/s 84–223 · low confidence	Radeon Instinct MI350X AMD	288 GB	8,190 GB/s	Jan 2025	110.8 GB	Q8_0	Comfortable
139 tok/s 84–223 · low confidence	Radeon Instinct MI355X AMD	288 GB	8,190 GB/s	Jan 2025	110.8 GB	Q8_0	Comfortable
131 tok/s 78–209 · low confidence	H100 SXM5 94 GB NVIDIA	94 GB	3,360 GB/s	Mar 2023	73.1 GB	Q5_K_M	Tight
111 tok/s 67–178 · low confidence	Radeon Instinct MI300 AMD	128 GB	6,550 GB/s	Jan 2023	110.8 GB	Q8_0	Tight
108 tok/s 65–173 · low confidence	H100 SXM5 64 GB NVIDIA	64 GB	2,020 GB/s	Mar 2023	54.2 GB	IQ4_XS	Tight
107 tok/s 64–170 · low confidence	H200 NVL NVIDIA	141 GB	4,890 GB/s	Nov 2024	110.8 GB	Q8_0	Tight
107 tok/s 64–170 · low confidence	H200 SXM 141 GB NVIDIA	141 GB	4,890 GB/s	Nov 2024	110.8 GB	Q8_0	Tight
106 tok/s 64–170 · low confidence	H100 PCIe 96 GB NVIDIA	96 GB	3,360 GB/s	Mar 2023	85.6 GB	Q6_K	Tight
106 tok/s 64–170 · low confidence	H100 SXM5 96 GB NVIDIA	96 GB	3,360 GB/s	Mar 2023	85.6 GB	Q6_K	Tight
103 tok/s 62–164 · low confidence	A100 SXM4 80 GB NVIDIA	80 GB	2,040 GB/s	Nov 2020	60.5 GB	Q4_K_M	Tight
103 tok/s 62–164 · low confidence	A100X NVIDIA	80 GB	2,040 GB/s	Jun 2021	60.5 GB	Q4_K_M	Tight
103 tok/s 62–164 · low confidence	A800 SXM4 80 GB NVIDIA	80 GB	2,040 GB/s	Aug 2022	60.5 GB	Q4_K_M	Tight
103 tok/s 62–164 · low confidence	H100 CNX NVIDIA	80 GB	2,040 GB/s	Mar 2023	60.5 GB	Q4_K_M	Tight
103 tok/s 62–164 · low confidence	H100 PCIe 80 GB NVIDIA	80 GB	2,040 GB/s	Oct 2022	60.5 GB	Q4_K_M	Tight
103 tok/s 62–164 · low confidence	H800 PCIe 80 GB NVIDIA	80 GB	2,040 GB/s	Mar 2023	60.5 GB	Q4_K_M	Tight
102 tok/s 61–163 · low confidence	Radeon Instinct MI325X AMD	256 GB	6,000 GB/s	Oct 2024	110.8 GB	Q8_0	Comfortable
97.6 tok/s 59–156 · low confidence	A100 PCIe 80 GB NVIDIA	80 GB	1,940 GB/s	Jun 2021	60.5 GB	Q4_K_M	Tight
97.6 tok/s 59–156 · low confidence	A800 PCIe 80 GB NVIDIA	80 GB	1,940 GB/s	Nov 2022	60.5 GB	Q4_K_M	Tight
90.5 tok/s 54–145 · low confidence	Radeon Instinct MI300A AMD	128 GB	5,325 GB/s	Dec 2023	110.8 GB	Q8_0	Tight
90.5 tok/s 54–145 · low confidence	Radeon Instinct MI300X AMD	192 GB	5,325 GB/s	Dec 2023	110.8 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: Z.ai (Zhipu AI),Tsinghua University
Organisation type: Industry,Academia
Country: China
Published: 15 August 2025
Authors: GLM-V Team: Wenyi Hong, Wenmeng Yu, Xiaotao Gu, Guo Wang, Guobing Gan, Haomiao Tang, Jiale Cheng, Ji Qi, Junhui Ji, Lihang Pan, Shuaiqi Duan, Weihan Wang, Yan Wang, Yean Cheng, Zehai He, Zhe Su, Zhen Yang, Ziyang Pan, Aohan Zeng, Baoxu Wang, Bin Chen, Boyan Shi, Changyu Pang, Chenhui Zhang, Da Yin, Fan Yang, Guoqing Chen, Jiazheng Xu, Jiale Zhu, Jiali Chen, Jing Chen, Jinhao Chen, Jinghao Lin, Jin…

What it does

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

Domain: Multimodal, Vision, Language, Video
Task: Language modeling/generation, Question answering, Visual question answering, Image captioning, Video description, Table tasks, Character recognition (OCR)
Base model: GLM-4.5-Air

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: 108B
Training data: 2,013,265,900,000 tokens

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.

Training compute: 1.8 × 10²⁴ FLOP
How it was established: Operation counting
Fine-tuning compute: 1.4 × 10²³ 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: Unreleased
Hugging Face: zai-org

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: GLM-4.5V and GLM-4.1V-Thinking: Towards Versatile Multimodal Reasoning with Scalable Reinforcement Learning
Last updated: 28 November 2025

The extremes

The ten fastest GPUs that run GLM-4.5V

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 GLM-4.5V

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

Hardware requirements in practice

Minimum card

Radeon Instinct MI200

Memory needed

54.2 GB

Fastest

174 tok/s

GLM-4.5V reaches a parameter count of 108B. That puts it above consumer hardware, into the range where a card is bought for this purpose rather than repurposed for it. The number of cards we track that can hold it: 43.

The entry point is Radeon Instinct MI200, with a memory capacity of 64 GB, running it at a compression of IQ4_XS and producing around 68.4 tokens per second.

Top of the range is B200, generating roughly 174 tokens per second on the strength of a memory bandwidth of 8,000 GB/s.

About this model

GLM-4.5V was published by Z.ai (Zhipu AI),Tsinghua University, in the country recorded as China, during August 2025. The publishing organisation is categorised as industry,Academia.

It works in the domain of Multimodal, Vision, Language, Video, and is recorded as performing the task of language modeling/generation, Question answering, Visual question answering, Image captioning, Video description, Table tasks, Character recognition (OCR).

Its starting point was an existing base model, GLM-4.5-Air. That is why it shares the base model's general shape and size.

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 zai-org.

How fast it runs, and why

Across every card that can run it, the middle of the range sits at 97.6 tokens per second. Producing text faster than most people read it: 41 of them.

Mixture-of-experts routing is why the speeds here look high for the parameter count. Only a fraction is read per token, but the whole thing has to be loaded.

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.

Training and provenance

The training run consumed about 1.8 × 10²⁴ FLOP. That figure measures what producing the model cost, and has no bearing on how fast it answers.

It was trained on a corpus of about 2,013,265,900,000 tokens of text.

Step by step

How to choose a GPU for GLM-4.5V

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 able to hold GLM-4.5V, needing around 54.2 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 GLM-4.5V.
03

Decide how much compression you will accept

Compression is what makes a model fit smaller cards, at some cost in accuracy, 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

Sort by speed to see how cards rank for GLM-4.5V. It will not match a gaming ordering, because generation is bound by memory bandwidth. The card topping the list is B200, at 174 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 GLM-4.5V. 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 GLM-4.5V.

Answers

GLM-4.5V — common questions

GLM-4.5V— 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.

GLM-4.5V— how accurate are these speed estimates?

Every figure is derived from memory bandwidth and model size, not benchmarked, which is why each is published as a range rather than a single number. One example: 105–279 tok/s on B200. The same model and card vary by thirty to fifty per cent depending on the inference software and its version.

GLM-4.5V— what GPU do I need to run it?

The smallest card in our catalogue that holds it is Radeon Instinct MI200, with a memory capacity of 64 GB. It runs the model at a compression of IQ4_XS using about 54.2 GB, and produces roughly 68.4 tokens per second. The number of cards able to run it in total: 43.

GLM-4.5V— how fast is it on a GPU?

It depends on the card. The quickest we calculate is B200, at about 174 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: 41.

GLM-4.5V— how much VRAM does it need?

It needs about 54.2 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.

GLM-4.5V— 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.

GLM-4.5V— how many parameters does it have?

It has a parameter count of 108B. 108 total, 12 B 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.

GLM-4.5V— who created it?

It was published by Z.ai (Zhipu AI),Tsinghua University, based in China, an organisation categorised as industry,Academia.

GLM-4.5V— when was it released?

It was published in August 2025.

GLM-4.5V— what is it used for?

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

GLM-4.5V— where can I download it?

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

GLM-4.5V— how much compute was used to train it?

Training consumed around 1.8 × 10²⁴ FLOP. That measures what producing the model cost and says nothing about how quickly it answers once trained — inference speed comes from memory bandwidth, not from the training budget.

GLM-4.5V— 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 17.3 GB. Every figure here assumes the whole model is resident on the card.

GLM-4.5V— 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: 43. So a second card is rarely the answer here.

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