ChatGLM3-6B TPS calculator

Open weights Z.ai (Zhipu AI) 6B parameters October 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 · Q4_K_M · 28.8 tok/s

Fastest card

B200

565 tok/s · 180 GB

Which GPUs can run ChatGLM3-6B?

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
565 tok/s 339–904 · low confidence	B200 NVIDIA	180 GB	8,000 GB/s	Jan 2024	7.1 GB	Q8_0	Comfortable
565 tok/s 339–904 · low confidence	B300 NVIDIA	288 GB	8,000 GB/s	Sep 2025	7.1 GB	Q8_0	Comfortable
451 tok/s 271–721 · low confidence	Radeon Instinct MI350X AMD	288 GB	8,190 GB/s	Jan 2025	7.1 GB	Q8_0	Comfortable
451 tok/s 271–721 · low confidence	Radeon Instinct MI355X AMD	288 GB	8,190 GB/s	Jan 2025	7.1 GB	Q8_0	Comfortable
361 tok/s 216–577 · low confidence	Radeon Instinct MI300 AMD	128 GB	6,550 GB/s	Jan 2023	7.1 GB	Q8_0	Comfortable
345 tok/s 207–552 · low confidence	H200 NVL NVIDIA	141 GB	4,890 GB/s	Nov 2024	7.1 GB	Q8_0	Comfortable
345 tok/s 207–552 · low confidence	H200 SXM 141 GB NVIDIA	141 GB	4,890 GB/s	Nov 2024	7.1 GB	Q8_0	Comfortable
330 tok/s 198–529 · low confidence	Radeon Instinct MI325X AMD	256 GB	6,000 GB/s	Oct 2024	7.1 GB	Q8_0	Comfortable
293 tok/s 176–469 · low confidence	Radeon Instinct MI300A AMD	128 GB	5,325 GB/s	Dec 2023	7.1 GB	Q8_0	Comfortable
293 tok/s 176–469 · low confidence	Radeon Instinct MI300X AMD	192 GB	5,325 GB/s	Dec 2023	7.1 GB	Q8_0	Comfortable
293 tok/s 176–469 · low confidence	Radeon Instinct MI308X AMD	192 GB	5,325 GB/s	Dec 2023	7.1 GB	Q8_0	Comfortable
278 tok/s 167–445 · low confidence	H100 NVL 94 GB NVIDIA	94 GB	3,940 GB/s	Mar 2023	7.1 GB	Q8_0	Comfortable
237 tok/s 142–379 · low confidence	H100 PCIe 96 GB NVIDIA	96 GB	3,360 GB/s	Mar 2023	7.1 GB	Q8_0	Comfortable
237 tok/s 142–379 · low confidence	H100 SXM5 80 GB NVIDIA	80 GB	3,360 GB/s	Oct 2022	7.1 GB	Q8_0	Comfortable
237 tok/s 142–379 · low confidence	H100 SXM5 94 GB NVIDIA	94 GB	3,360 GB/s	Mar 2023	7.1 GB	Q8_0	Comfortable
237 tok/s 142–379 · low confidence	H100 SXM5 96 GB NVIDIA	96 GB	3,360 GB/s	Mar 2023	7.1 GB	Q8_0	Comfortable
237 tok/s 142–379 · low confidence	H800 SXM5 NVIDIA	80 GB	3,360 GB/s	Mar 2023	7.1 GB	Q8_0	Comfortable
181 tok/s 108–289 · low confidence	Radeon Instinct MI250 AMD	128 GB	3,280 GB/s	Nov 2021	7.1 GB	Q8_0	Comfortable
181 tok/s 108–289 · low confidence	Radeon Instinct MI250X AMD	128 GB	3,280 GB/s	Nov 2021	7.1 GB	Q8_0	Comfortable
150 tok/s 90–241 · low confidence	Data Center GPU Max 1550 Intel	128 GB	3,280 GB/s	Jan 2023	7.1 GB	Q8_0	Comfortable
147 tok/s 88–236 · low confidence	Data Center GPU Max Subsystem Intel	128 GB	3,210 GB/s	Jan 2023	7.1 GB	Q8_0	Comfortable
144 tok/s 86–230 · low confidence	A100 SXM4 80 GB NVIDIA	80 GB	2,040 GB/s	Nov 2020	7.1 GB	Q8_0	Comfortable
144 tok/s 86–230 · low confidence	A100X NVIDIA	80 GB	2,040 GB/s	Jun 2021	7.1 GB	Q8_0	Comfortable
144 tok/s 86–230 · low confidence	A800 SXM4 80 GB NVIDIA	80 GB	2,040 GB/s	Aug 2022	7.1 GB	Q8_0	Comfortable
144 tok/s 86–230 · low confidence	H100 CNX NVIDIA	80 GB	2,040 GB/s	Mar 2023	7.1 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: Z.ai (Zhipu AI)
Organisation type: Industry
Country: China
Published: 27 October 2023
Authors: Aohan Zeng, Bin Xu, Bowen Wang, Chenhui Zhang, Da Yin, Diego Rojas, Guanyu Feng, Hanlin Zhao, Hanyu Lai, Hao Yu, Hongning Wang, Jiadai Sun, Jiajie Zhang, Jiale Cheng, Jiayi Gui, Jie Tang, Jing Zhang, Juanzi Li, Lei Zhao, Lindong Wu, Lucen Zhong, Mingdao Liu, Minlie Huang, Peng Zhang, Qinkai Zheng, Rui Lu, Shuaiqi Duan, Shudan Zhang, Shulin Cao, Shuxun Yang, Weng Lam Tam, Wenyi Zhao, Xiao Liu, Xiao…

What it does

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

Domain: Multimodal, Language, Vision
Task: Chat, Visual question answering, Code generation

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: 6B
Training data: 1,400,000,000,000 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: 5 × 10²² FLOP
How it was established: Operation counting

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 (restricted use)
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
Why it is tracked: SOTA improvement
Record confidence: Speculative

Sources

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

Reference: Zhipu AI launches third-generation base model
Last updated: 28 November 2025

The extremes

The ten fastest GPUs that run ChatGLM3-6B

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 ChatGLM3-6B

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

Tesla K20c

Memory needed

4.3 GB

Fastest

565 tok/s

ChatGLM3-6B is small enough at 6B parameters that hardware is rarely the obstacle — 589 of the cards we track can run it, including cards several years old.

The smallest card that holds it is the Tesla K20c with 5 GB, running it at Q4_K_M and producing around 28.8 tokens per second.

A B200 is the fastest we calculate for it: about 565 tokens per second, from 8,000 GB/s of memory bandwidth.

Background

ChatGLM3-6B was published by Z.ai (Zhipu AI), in China, in October 2023. industry is the category the publisher falls under.

It works in Multimodal, Language, Vision, and is recorded as doing chat, Visual question answering, Code generation.

Because its weights were released, nothing about running it depends on a provider staying available — it is yours once downloaded.

Reading the throughput figures

The median result is around 25.4 tokens per second; 539 cards produce text faster than most people read it.

Every weight participates in every token here, so bandwidth is the whole story: the ranking below is effectively a ranking of memory throughput.

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 5 × 10²² FLOP. That figure describes the cost of creating it and has no bearing on how quickly it generates text.

The training set ran to roughly 1,400,000,000,000 tokens.

Its inclusion criterion is sOTA improvement.

Step by step

How to choose a GPU for ChatGLM3-6B

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

Look at what ChatGLM3-6B actually needs — around 4.3 GB at Q4_K_M. No amount of processing power compensates for a card that cannot hold it.
02

Match the context to your actual use

Longer conversations cost memory on top of what the weights need. Move the slider to your real working length before trusting any row for ChatGLM3-6B.
03

Choose how far you will compress it

Compression is what makes ChatGLM3-6B fit smaller cards, at some cost in accuracy — Q4_K_M on the smallest card that fits. A minimum quality removes the ones that go too far.
04

Rank by throughput rather than spec sheet

Ranking by tokens per second for ChatGLM3-6B follows memory bandwidth, not core counts, which is why the B200 tops it at 565 tok/s.
05

Read the fit column last

Tight means ChatGLM3-6B loads and works, with no room to raise the context later. Comfortable means you can. The difference matters more than a few tokens per second.
06

Check the card from the other side

Every card name links to its own page, which runs the same calculation across the whole model catalogue. Worth a look before buying for ChatGLM3-6B alone — a card is usually bought for more than one model.

Answers

ChatGLM3-6B — common questions

How much compute was used to train ChatGLM3-6B?

Around 5 × 10²² 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.

Can I run ChatGLM3-6B 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 ChatGLM3-6B is rarely worth using — the nearest miss we calculate is short by 0.7 GB. Every figure here assumes the whole model is on the card.

Would two GPUs run ChatGLM3-6B faster?

A second card roughly doubles the memory available but not the generation rate. With 589 cards already able to run ChatGLM3-6B alone, the case for pairing is weak.

Why does the quantisation differ between cards for ChatGLM3-6B?

Each card is shown running the least-compressed copy it can hold, and ChatGLM3-6B appears at 3 different compression levels across the cards that fit it. Bigger cards get the more accurate version.

How accurate are these ChatGLM3-6B speed estimates?

These are estimates with real error bars. The fastest result here, 339–904 tok/s on the B200, could reasonably land anywhere in its published range depending on which runtime you use.

What GPU do I need to run ChatGLM3-6B?

The smallest card in our catalogue that holds ChatGLM3-6B is the Tesla K20c, with 5 GB of memory. It runs the model at Q4_K_M using about 4.3 GB, and produces roughly 28.8 tokens per second. 589 cards in total can run it.

How fast is ChatGLM3-6B on a GPU?

It depends on the card. The quickest we calculate is a B200 at about 565 tokens per second; the slowest that still runs it manages considerably less. Reading speed is around ten tokens per second, and 539 of the cards that can run ChatGLM3-6B clear that.

How much VRAM does ChatGLM3-6B need?

About 4.3 GB at Q4_K_M compression, 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.

Can I run ChatGLM3-6B on a 8 GB GPU?

Yes. A CMP 170HX 8 GB with 8 GB runs it at Q8_0, using about 7.1 GB and generating roughly 105 tokens per second — a tight fit.

Can I run ChatGLM3-6B on a 12 GB GPU?

Yes. A GeForce RTX 3080 Ti with 12 GB runs it at Q8_0, using about 7.1 GB and generating roughly 64.4 tokens per second — a comfortable fit.

Can I run ChatGLM3-6B on a 16 GB GPU?

Yes. A Tesla V100 SXM2 16 GB with 16 GB runs it at Q8_0, using about 7.1 GB and generating roughly 79.8 tokens per second — a comfortable fit.

Can I run ChatGLM3-6B on a 24 GB GPU?

Yes. A GeForce RTX 5090 D V2 with 24 GB runs it at Q8_0, using about 7.1 GB and generating roughly 94.6 tokens per second — a comfortable fit.

Is ChatGLM3-6B open source?

Its weights are published, so ChatGLM3-6B 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.

How many parameters does ChatGLM3-6B have?

ChatGLM3-6B has 6B parameters. 6B from https://arxiv.org/abs/2406.12793. 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.

Who created ChatGLM3-6B?

ChatGLM3-6B was published by Z.ai (Zhipu AI), based in China, categorised as industry.

When was ChatGLM3-6B released?

ChatGLM3-6B was published in October 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.

What is ChatGLM3-6B used for?

ChatGLM3-6B works in Multimodal, Language, Vision, and is recorded as handling chat, Visual question answering, Code generation. Models frequently carry more than one of each, and the tags describe purpose rather than capability limits.

Where can I download ChatGLM3-6B?

The weights for ChatGLM3-6B are published, though we do not hold a repository link for it. This site calculates hardware requirements rather than hosting model files.

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