Kimi Linear TPS calculator

Open weights Moonshot 48B parameters October 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

93 cards that can run it

818 cards we hold specifications for

Smallest card that fits

Radeon PRO V710

28 GB · Q3_K_M · 9.4 tok/s

Fastest card

B200

70.6 tok/s · 180 GB

Which GPUs can run Kimi Linear?

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

93 cards match

Calculating

					Needs	Quantisation	Fit
70.6 tok/s 42–113 · low confidence	B200 NVIDIA	180 GB	8,000 GB/s	Jan 2024	52.1 GB	Q8_0	Comfortable
70.6 tok/s 42–113 · low confidence	B300 NVIDIA	288 GB	8,000 GB/s	Sep 2025	52.1 GB	Q8_0	Comfortable
56.4 tok/s 34–90 · low confidence	Radeon Instinct MI350X AMD	288 GB	8,190 GB/s	Jan 2025	52.1 GB	Q8_0	Comfortable
56.4 tok/s 34–90 · low confidence	Radeon Instinct MI355X AMD	288 GB	8,190 GB/s	Jan 2025	52.1 GB	Q8_0	Comfortable
45.1 tok/s 27–72 · low confidence	Radeon Instinct MI300 AMD	128 GB	6,550 GB/s	Jan 2023	52.1 GB	Q8_0	Comfortable
43.2 tok/s 26–69 · low confidence	H200 NVL NVIDIA	141 GB	4,890 GB/s	Nov 2024	52.1 GB	Q8_0	Comfortable
43.2 tok/s 26–69 · low confidence	H200 SXM 141 GB NVIDIA	141 GB	4,890 GB/s	Nov 2024	52.1 GB	Q8_0	Comfortable
41.3 tok/s 25–66 · low confidence	Radeon Instinct MI325X AMD	256 GB	6,000 GB/s	Oct 2024	52.1 GB	Q8_0	Comfortable
40.5 tok/s 24–65 · low confidence	DRIVE A100 PROD NVIDIA	32 GB	1,870 GB/s	May 2020	26.9 GB	IQ4_XS	Tight
40.5 tok/s 24–65 · low confidence	GRID A100A NVIDIA	32 GB	1,870 GB/s	May 2020	26.9 GB	IQ4_XS	Tight
38.8 tok/s 23–62 · low confidence	GeForce RTX 5090 NVIDIA	32 GB	1,790 GB/s	Jan 2025	26.9 GB	IQ4_XS	Tight
38.8 tok/s 23–62 · low confidence	GeForce RTX 5090 D NVIDIA	32 GB	1,790 GB/s	Jan 2025	26.9 GB	IQ4_XS	Tight
36.7 tok/s 22–59 · low confidence	Radeon Instinct MI300A AMD	128 GB	5,325 GB/s	Dec 2023	52.1 GB	Q8_0	Comfortable
36.7 tok/s 22–59 · low confidence	Radeon Instinct MI300X AMD	192 GB	5,325 GB/s	Dec 2023	52.1 GB	Q8_0	Comfortable
36.7 tok/s 22–59 · low confidence	Radeon Instinct MI308X AMD	192 GB	5,325 GB/s	Dec 2023	52.1 GB	Q8_0	Comfortable
34.8 tok/s 21–56 · low confidence	H100 NVL 94 GB NVIDIA	94 GB	3,940 GB/s	Mar 2023	52.1 GB	Q8_0	Comfortable
29.7 tok/s 18–47 · low confidence	H100 PCIe 96 GB NVIDIA	96 GB	3,360 GB/s	Mar 2023	52.1 GB	Q8_0	Comfortable
29.7 tok/s 18–47 · low confidence	H100 SXM5 80 GB NVIDIA	80 GB	3,360 GB/s	Oct 2022	52.1 GB	Q8_0	Comfortable
29.7 tok/s 18–47 · low confidence	H100 SXM5 94 GB NVIDIA	94 GB	3,360 GB/s	Mar 2023	52.1 GB	Q8_0	Comfortable
29.7 tok/s 18–47 · low confidence	H100 SXM5 96 GB NVIDIA	96 GB	3,360 GB/s	Mar 2023	52.1 GB	Q8_0	Comfortable
29.7 tok/s 18–47 · low confidence	H800 SXM5 NVIDIA	80 GB	3,360 GB/s	Mar 2023	52.1 GB	Q8_0	Comfortable
24.6 tok/s 15–39 · low confidence	A100 PCIe 40 GB NVIDIA	40 GB	1,560 GB/s	Jun 2020	35.3 GB	Q5_K_M	Tight
24.6 tok/s 15–39 · low confidence	A100 SXM4 40 GB NVIDIA	40 GB	1,560 GB/s	May 2020	35.3 GB	Q5_K_M	Tight
24.6 tok/s 15–39 · low confidence	A800 PCIe 40 GB NVIDIA	40 GB	1,560 GB/s	Nov 2022	35.3 GB	Q5_K_M	Tight
24.5 tok/s 15–39 · low confidence	Tesla PG500-216 NVIDIA	32 GB	1,130 GB/s	Nov 2019	26.9 GB	IQ4_XS	Tight

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: Moonshot
Organisation type: Industry
Country: China
Published: 30 October 2025

What it does

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

Domain: Language
Task: Language modeling/generation, Question answering, Mathematical reasoning

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: 48B
Training data: 5,700,000,000,000 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: 1 × 10²³ 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 (unrestricted)
Training code: Unreleased
Hugging Face: moonshotai

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: KIMI LINEAR: AN EXPRESSIVE, EFFICIENT ATTENTION ARCHITECTURE
Last updated: 28 November 2025

The extremes

The ten fastest GPUs that run Kimi Linear

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 Kimi Linear

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 you need to run it

Minimum card

Radeon PRO V710

Memory needed

24.2 GB

Fastest

70.6 tok/s

Kimi Linear reaches a parameter count of 48B. 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: 93.

The least hardware that works is Radeon PRO V710, with a memory capacity of 28 GB, running it at a compression of Q3_K_M and producing around 9.4 tokens per second.

The quickest result comes from B200, generating roughly 70.6 tokens per second on the strength of a memory bandwidth of 8,000 GB/s.

What this model is

Kimi Linear was published by Moonshot, in the country recorded as China, during October 2025. The category the publisher falls under is industry.

It works in the domain of Language, and is recorded as performing the task of language modeling/generation, Question answering, Mathematical reasoning.

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. On Hugging Face it is published under the organisation moonshotai.

What decides the speed

Across every card that can run it, the middle of the range sits at 17.8 tokens per second. Producing text faster than most people read it: 71 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.

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.

Training and provenance

Training it took a computation budget of roughly 1 × 10²³ FLOP. That figure measures what producing the model cost, and has no bearing on how fast it answers.

Training consumed a corpus of around 5,700,000,000,000 tokens of text.

Step by step

How to choose a GPU for Kimi Linear

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

Every card here has been checked against Kimi Linear, needing around 24.2 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 Kimi Linear.
03

Choose how far you will compress it

Each card runs the least-compressed copy it can hold, 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 Kimi Linear. It will not match a gaming ordering, because generation is bound by memory bandwidth. The card topping the list is B200, at 70.6 tok/s.
05

Read the fit column last

A tight fit runs, but leaves nothing spare for a longer conversation, in the case of Kimi Linear. 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

Following a card through to its own page shows every other model it can hold, which is the question that follows once you have settled on Kimi Linear.

Answers

Kimi Linear — common questions

Kimi Linear— 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 8.1 GB. Every figure here assumes the whole model is resident on the card.

Kimi Linear— 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: 93. So a second card is rarely the answer here.

Kimi Linear— why does the quantisation differ between cards?

Because capacity varies, so does how hard it has to be squeezed. The number of distinct levels in the table above: 5. Bigger cards get the more accurate version, and the quality floor above pins the comparison to one level.

Kimi Linear— how accurate are these speed estimates?

They are calculated from specifications rather than measured, and each carries a range. One example: 42–113 tok/s on B200. The same model and card vary by thirty to fifty per cent depending on the inference software and its version.

Kimi Linear— what GPU do I need to run it?

The smallest card in our catalogue that holds it is Radeon PRO V710, with a memory capacity of 28 GB. It runs the model at a compression of Q3_K_M using about 24.2 GB, and produces roughly 9.4 tokens per second. The number of cards able to run it in total: 93.

Kimi Linear— how fast is it on a GPU?

It depends on the card. The quickest we calculate is B200, at about 70.6 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: 71.

Kimi Linear— how much VRAM does it need?

It needs about 24.2 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.

Kimi Linear— 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.

Kimi Linear— how many parameters does it have?

It has a parameter count of 48B. 3B activated parameters and 48B total 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.

Kimi Linear— who created it?

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

Kimi Linear— when was it released?

It was published in October 2025.

Kimi Linear— what is it used for?

It works in the domain of Language, and is recorded as handling the task of language modeling/generation, Question answering, Mathematical reasoning. Models frequently carry more than one of each, and the tags describe purpose rather than capability limits.

Kimi Linear— where can I download it?

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

Kimi Linear— how much compute was used to train it?

Training consumed around 1 × 10²³ 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.

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