GLM-130B TPS calculator

Open weights Tsinghua University 130B parameters August 2022

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

38 cards that can run it

818 cards we hold specifications for

Smallest card that fits

RTX PRO 5000 72 GB Blackwell

72 GB · Q3_K_M · 11.8 tok/s

Fastest card

H100 NVL 94 GB

29.6 tok/s · 94 GB

Which GPUs can run GLM-130B?

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

38 cards match

Calculating

					Needs	Quantisation	Fit
29.6 tok/s 18–47 · low confidence	H100 NVL 94 GB NVIDIA	94 GB	3,940 GB/s	Mar 2023	79.3 GB	Q4_K_M	Tight
26.9 tok/s 16–43 · low confidence	H100 SXM5 80 GB NVIDIA	80 GB	3,360 GB/s	Oct 2022	71.8 GB	IQ4_XS	Tight
26.9 tok/s 16–43 · low confidence	H800 SXM5 NVIDIA	80 GB	3,360 GB/s	Mar 2023	71.8 GB	IQ4_XS	Tight
26.1 tok/s 16–42 · low confidence	B200 NVIDIA	180 GB	8,000 GB/s	Jan 2024	139.9 GB	Q8_0	Tight
26.1 tok/s 16–42 · low confidence	B300 NVIDIA	288 GB	8,000 GB/s	Sep 2025	139.9 GB	Q8_0	Comfortable
25.3 tok/s 15–40 · low confidence	H100 PCIe 96 GB NVIDIA	96 GB	3,360 GB/s	Mar 2023	79.3 GB	Q4_K_M	Tight
25.3 tok/s 15–40 · low confidence	H100 SXM5 94 GB NVIDIA	94 GB	3,360 GB/s	Mar 2023	79.3 GB	Q4_K_M	Tight
25.3 tok/s 15–40 · low confidence	H100 SXM5 96 GB NVIDIA	96 GB	3,360 GB/s	Mar 2023	79.3 GB	Q4_K_M	Tight
24.2 tok/s 15–39 · low confidence	Radeon Instinct MI300 AMD	128 GB	6,550 GB/s	Jan 2023	109.6 GB	Q6_K	Tight
23.2 tok/s 14–37 · low confidence	H200 NVL NVIDIA	141 GB	4,890 GB/s	Nov 2024	109.6 GB	Q6_K	Tight
23.2 tok/s 14–37 · low confidence	H200 SXM 141 GB NVIDIA	141 GB	4,890 GB/s	Nov 2024	109.6 GB	Q6_K	Tight
20.8 tok/s 12–33 · low confidence	Radeon Instinct MI350X AMD	288 GB	8,190 GB/s	Jan 2025	139.9 GB	Q8_0	Comfortable
20.8 tok/s 12–33 · low confidence	Radeon Instinct MI355X AMD	288 GB	8,190 GB/s	Jan 2025	139.9 GB	Q8_0	Comfortable
19.7 tok/s 12–31 · low confidence	Radeon Instinct MI300A AMD	128 GB	5,325 GB/s	Dec 2023	109.6 GB	Q6_K	Tight
16.3 tok/s 10–26 · low confidence	A100 SXM4 80 GB NVIDIA	80 GB	2,040 GB/s	Nov 2020	71.8 GB	IQ4_XS	Tight
16.3 tok/s 10–26 · low confidence	A100X NVIDIA	80 GB	2,040 GB/s	Jun 2021	71.8 GB	IQ4_XS	Tight
16.3 tok/s 10–26 · low confidence	A800 SXM4 80 GB NVIDIA	80 GB	2,040 GB/s	Aug 2022	71.8 GB	IQ4_XS	Tight
16.3 tok/s 10–26 · low confidence	H100 CNX NVIDIA	80 GB	2,040 GB/s	Mar 2023	71.8 GB	IQ4_XS	Tight
16.3 tok/s 10–26 · low confidence	H100 PCIe 80 GB NVIDIA	80 GB	2,040 GB/s	Oct 2022	71.8 GB	IQ4_XS	Tight
16.3 tok/s 10–26 · low confidence	H800 PCIe 80 GB NVIDIA	80 GB	2,040 GB/s	Mar 2023	71.8 GB	IQ4_XS	Tight
15.5 tok/s 9–25 · low confidence	A100 PCIe 80 GB NVIDIA	80 GB	1,940 GB/s	Jun 2021	71.8 GB	IQ4_XS	Tight
15.5 tok/s 9–25 · low confidence	A800 PCIe 80 GB NVIDIA	80 GB	1,940 GB/s	Nov 2022	71.8 GB	IQ4_XS	Tight
15.3 tok/s 9–24 · low confidence	Radeon Instinct MI325X AMD	256 GB	6,000 GB/s	Oct 2024	139.9 GB	Q8_0	Comfortable
13.5 tok/s 8–22 · low confidence	Radeon Instinct MI300X AMD	192 GB	5,325 GB/s	Dec 2023	139.9 GB	Q8_0	Comfortable
13.5 tok/s 8–22 · low confidence	Radeon Instinct MI308X AMD	192 GB	5,325 GB/s	Dec 2023	139.9 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: Tsinghua University
Organisation type: Academia
Country: China
Published: 4 August 2022
Authors: Aohan Zeng, Xiao Liu, Zhengxiao Du, Zihan Wang, Hanyu Lai, Ming Ding, Zhuoyi Yang, Yifan Xu, Wendi Zheng, Xiao Xia, Weng Lam Tam, Zixuan Ma, Yufei Xue, Jidong Zhai, Wenguang Chen, Peng Zhang, Yuxiao Dong, Jie Tang

What it does

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

Domain: Language
Task: Language modeling/generation, Translation
Numerical format: FP16

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: 130B
Training data: 152,000,000,000 tokens
Epochs: 1
Batch size: 8,650,752

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: 3.5 × 10²³ FLOP
How it was established: Operation counting,Hardware

The training run

What it physically took to train: which chips, how many, for how long, and what that drew from the wall.

Training hardware: NVIDIA A100 SXM4 40 GB
Chips used: 768
Chip-hours: 1,105,920
Wall-clock time: 1,440 hours (60 days)
Hardware utilisation: MFU 32.5% · HFU 43.3%
Power draw: 615.7 kW
Compute cost: $820,297

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

How it is classified

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

Foundation model: Yes
Likely above 10²³ FLOP: Yes
Why it is tracked: SOTA improvement
Record confidence: Confident
Citations: 1,264
Benchmark data: GLM-130B

Sources

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

Reference: GLM-130B: An Open Bilingual Pre-trained Model
Last updated: 25 May 2026

The extremes

The ten fastest GPUs that run GLM-130B

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-130B

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 PRO 5000 72 GB Blackwell

Memory needed

64.2 GB

Fastest

29.6 tok/s

GLM-130B reaches a parameter count of 130B. 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: 38.

At the low end it is handled by RTX PRO 5000 72 GB Blackwell, with a memory capacity of 72 GB, running it at a compression of Q3_K_M and producing around 11.8 tokens per second.

The quickest result comes from H100 NVL 94 GB, generating roughly 29.6 tokens per second on the strength of a memory bandwidth of 3,940 GB/s.

About this model

GLM-130B was published by Tsinghua University, in the country recorded as China, during August 2022. The category the publisher falls under is academia.

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

The weights being open is what puts this page in the calculator rather than only in the catalogue: it is a model you can actually hold.

How fast it runs, and why

Half the cards that hold it manage more than 16.3 tokens per second. Clearing the ten tokens per second that roughly matches reading speed: 35 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.

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.

What went into building it

The training run consumed about 3.5 × 10²³ FLOP, on hardware recorded as NVIDIA A100 SXM4 40 GB. That figure measures what producing the model cost, and has no bearing on how fast it answers.

Training consumed a corpus of around 152,000,000,000 tokens of text.

It is tracked in the underlying dataset for one reason in particular: sOTA improvement.

Step by step

How to choose a GPU for GLM-130B

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

Every card here has been checked against GLM-130B, needing around 64.2 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

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-130B.
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 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

Rank by throughput rather than spec sheet

The speed ordering is effectively an ordering by memory bandwidth, for GLM-130B. It will not match a gaming ordering, because generation is bound by memory bandwidth. The card topping the list is H100 NVL 94 GB, at 29.6 tok/s.
05

Read the fit column last

A tight fit runs, but leaves nothing spare for a longer conversation, in the case of GLM-130B. 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 GLM-130B.

Answers

GLM-130B — common questions

GLM-130B— how many parameters does it have?

It has a parameter count of 130B. Dense model. 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-130B— who created it?

It was published by Tsinghua University, based in China, an organisation categorised as academia.

GLM-130B— when was it released?

It was published in August 2022. Capability per parameter has improved considerably since, so a newer model of the same size is often the better use of the same hardware.

GLM-130B— what is it used for?

It works in the domain of Language, and is recorded as handling the task of language modeling/generation, Translation. 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.

GLM-130B— 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.

GLM-130B— how much compute was used to train it?

Training consumed around 3.5 × 10²³ FLOP, on hardware recorded as NVIDIA A100 SXM4 40 GB. 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-130B— 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 21.7 GB. Every figure here assumes the whole model is resident on the card.

GLM-130B— would two GPUs run it faster?

Two cards buy memory rather than speed, which matters only if one card cannot hold it. The number that can: 38. So a second card is rarely the answer here.

GLM-130B— 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-130B— 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: 18–47 tok/s on H100 NVL 94 GB. The same model and card vary by thirty to fifty per cent depending on the inference software and its version.

GLM-130B— what GPU do I need to run it?

The smallest card in our catalogue that holds it is RTX PRO 5000 72 GB Blackwell, with a memory capacity of 72 GB. It runs the model at a compression of Q3_K_M using about 64.2 GB, and produces roughly 11.8 tokens per second. The number of cards able to run it in total: 38.

GLM-130B— how fast is it on a GPU?

It depends on the card. The quickest we calculate is H100 NVL 94 GB, at about 29.6 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: 35.

GLM-130B— how much VRAM does it need?

It needs about 64.2 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.

GLM-130B— 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 25 May 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