Deephermes 3 Llama 3 8B Preview TPS calculator

Open weights Nous Research 8B parameters February 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

582 cards that can run it

818 cards we hold specifications for

Smallest card that fits

Quadro 6000

6 GB · Q3_K_M · 17.4 tok/s

Fastest card

B200

424 tok/s · 180 GB

Which GPUs can run Deephermes 3 Llama 3 8B Preview?

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

582 cards match

Calculating

					Needs	Quantisation	Fit
424 tok/s 360–508	B200 NVIDIA	180 GB	8,000 GB/s	Jan 2024	9.8 GB	Q8_0	Comfortable
424 tok/s 360–508	B300 NVIDIA	288 GB	8,000 GB/s	Sep 2025	9.8 GB	Q8_0	Comfortable
338 tok/s 203–541 · low confidence	Radeon Instinct MI350X AMD	288 GB	8,190 GB/s	Jan 2025	9.8 GB	Q8_0	Comfortable
338 tok/s 203–541 · low confidence	Radeon Instinct MI355X AMD	288 GB	8,190 GB/s	Jan 2025	9.8 GB	Q8_0	Comfortable
270 tok/s 162–433 · low confidence	Radeon Instinct MI300 AMD	128 GB	6,550 GB/s	Jan 2023	9.8 GB	Q8_0	Comfortable
259 tok/s 220–311	H200 NVL NVIDIA	141 GB	4,890 GB/s	Nov 2024	9.8 GB	Q8_0	Comfortable
259 tok/s 220–311	H200 SXM 141 GB NVIDIA	141 GB	4,890 GB/s	Nov 2024	9.8 GB	Q8_0	Comfortable
248 tok/s 149–396 · low confidence	Radeon Instinct MI325X AMD	256 GB	6,000 GB/s	Oct 2024	9.8 GB	Q8_0	Comfortable
220 tok/s 132–352 · low confidence	Radeon Instinct MI300A AMD	128 GB	5,325 GB/s	Dec 2023	9.8 GB	Q8_0	Comfortable
220 tok/s 132–352 · low confidence	Radeon Instinct MI300X AMD	192 GB	5,325 GB/s	Dec 2023	9.8 GB	Q8_0	Comfortable
220 tok/s 132–352 · low confidence	Radeon Instinct MI308X AMD	192 GB	5,325 GB/s	Dec 2023	9.8 GB	Q8_0	Comfortable
209 tok/s 177–250	H100 NVL 94 GB NVIDIA	94 GB	3,940 GB/s	Mar 2023	9.8 GB	Q8_0	Comfortable
178 tok/s 151–213	H100 PCIe 96 GB NVIDIA	96 GB	3,360 GB/s	Mar 2023	9.8 GB	Q8_0	Comfortable
178 tok/s 151–213	H100 SXM5 80 GB NVIDIA	80 GB	3,360 GB/s	Oct 2022	9.8 GB	Q8_0	Comfortable
178 tok/s 151–213	H100 SXM5 94 GB NVIDIA	94 GB	3,360 GB/s	Mar 2023	9.8 GB	Q8_0	Comfortable
178 tok/s 151–213	H100 SXM5 96 GB NVIDIA	96 GB	3,360 GB/s	Mar 2023	9.8 GB	Q8_0	Comfortable
178 tok/s 151–213	H800 SXM5 NVIDIA	80 GB	3,360 GB/s	Mar 2023	9.8 GB	Q8_0	Comfortable
141 tok/s 120–169	CMP 170HX 8 GB NVIDIA	8 GB	1,490 GB/s	Sep 2021	7.0 GB	Q5_K_M	Tight
135 tok/s 81–217 · low confidence	Radeon Instinct MI250 AMD	128 GB	3,280 GB/s	Nov 2021	9.8 GB	Q8_0	Comfortable
135 tok/s 81–217 · low confidence	Radeon Instinct MI250X AMD	128 GB	3,280 GB/s	Nov 2021	9.8 GB	Q8_0	Comfortable
120 tok/s 102–144	CMP 170HX 10 GB NVIDIA	10 GB	1,560 GB/s	Sep 2021	7.9 GB	Q6_K	Tight
113 tok/s 68–181 · low confidence	Data Center GPU Max 1550 Intel	128 GB	3,280 GB/s	Jan 2023	9.8 GB	Q8_0	Comfortable
110 tok/s 66–177 · low confidence	Data Center GPU Max Subsystem Intel	128 GB	3,210 GB/s	Jan 2023	9.8 GB	Q8_0	Comfortable
108 tok/s 92–130	A100 SXM4 80 GB NVIDIA	80 GB	2,040 GB/s	Nov 2020	9.8 GB	Q8_0	Comfortable
108 tok/s 92–130	A100X NVIDIA	80 GB	2,040 GB/s	Jun 2021	9.8 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: Nous Research
Organisation type: Industry
Country: United States of America
Published: 14 February 2025
Authors: Teknium, Roger Jin, Chen Guang, Jai Suphavadeeprasit, Jeffrey Quesnelle

What it does

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

Domain: Language
Task: Mathematical reasoning, Quantitative reasoning, Language modeling/generation, Question answering
Base model: Llama 3.1-8B

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: 8B
Training data: 1,500,000,000 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.

How it was established: Operation counting
Fine-tuning compute: 7.2 × 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 (restricted use)
Training code: Unreleased
Hugging Face: NousResearch

How it is classified

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

Record confidence: Speculative

Sources

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

Reference: ‘Personalized, unrestricted’ AI lab Nous Research launches first toggle-on reasoning model: DeepHermes-3
Last updated: 28 November 2025

The extremes

The ten fastest GPUs that run Deephermes 3 Llama 3 8B Preview

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 Deephermes 3 Llama 3 8B Preview

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

Quadro 6000

Memory needed

5.1 GB

Fastest

424 tok/s

Deephermes 3 Llama 3 8B Preview reaches a parameter count of 8B. 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: 582.

At the low end it is handled by Quadro 6000, with a memory capacity of 6 GB, running it at a compression of Q3_K_M and producing around 17.4 tokens per second.

Top of the range is B200, generating roughly 424 tokens per second on the strength of a memory bandwidth of 8,000 GB/s.

Where it came from

Deephermes 3 Llama 3 8B Preview was published by Nous Research, in the country recorded as United States of America, during February 2025. The category the publisher falls under is industry.

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

Its starting point was an existing base model, Llama 3.1-8B. 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 NousResearch.

Understanding the speeds

Half the cards that hold it manage more than 23.8 tokens per second. Producing text faster than most people read it: 551 of them.

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

Because the architecture is recorded, the memory column is derived rather than estimated.

What went into building it

Training consumed a corpus of around 1,500,000,000 tokens of text.

Step by step

How to choose a GPU for Deephermes 3 Llama 3 8B Preview

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

The table lists every card able to hold Deephermes 3 Llama 3 8B Preview, needing around 5.1 GB at a compression of Q3_K_M. No amount of processing power compensates for a card that cannot hold it.
02

Decide how long your conversations run

Longer conversations cost memory on top of what the weights need. Move the slider to your real working length before trusting any row for Deephermes 3 Llama 3 8B Preview.
03

Set a quality floor

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

Ranking by tokens per second follows memory bandwidth rather than core counts, for Deephermes 3 Llama 3 8B Preview. It will not match a gaming ordering, because generation is bound by memory bandwidth. The card topping the list is B200, at 424 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 Deephermes 3 Llama 3 8B Preview. 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

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 Deephermes 3 Llama 3 8B Preview.

Answers

Deephermes 3 Llama 3 8B Preview — common questions

Deephermes 3 Llama 3 8B Preview— how accurate are these speed estimates?

They are calculated from specifications rather than measured, and each carries a range. One example: 360–508 tok/s on B200. The same model and card vary by thirty to fifty per cent depending on the inference software and its version.

Deephermes 3 Llama 3 8B Preview— what GPU do I need to run it?

The smallest card in our catalogue that holds it is Quadro 6000, with a memory capacity of 6 GB. It runs the model at a compression of Q3_K_M using about 5.1 GB, and produces roughly 17.4 tokens per second. The number of cards able to run it in total: 582.

Deephermes 3 Llama 3 8B Preview— how fast is it on a GPU?

It depends on the card. The quickest we calculate is B200, at about 424 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: 551.

Deephermes 3 Llama 3 8B Preview— how much VRAM does it need?

It needs about 5.1 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.

Deephermes 3 Llama 3 8B Preview— 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 Q5_K_M, using about 7.0 GB and generating roughly 141 tokens per second. The fit is tight.

Deephermes 3 Llama 3 8B Preview— 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 9.8 GB and generating roughly 48.3 tokens per second. The fit is tight.

Deephermes 3 Llama 3 8B Preview— 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 9.8 GB and generating roughly 59.8 tokens per second. The fit is comfortable.

Deephermes 3 Llama 3 8B Preview— 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 9.8 GB and generating roughly 70.9 tokens per second. The fit is comfortable.

Deephermes 3 Llama 3 8B Preview— 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.

Deephermes 3 Llama 3 8B Preview— how many parameters does it have?

It has a parameter count of 8B. 8B. 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.

Deephermes 3 Llama 3 8B Preview— who created it?

It was published by Nous Research, based in United States of America, an organisation categorised as industry.

Deephermes 3 Llama 3 8B Preview— when was it released?

It was published in February 2025.

Deephermes 3 Llama 3 8B Preview— what is it used for?

It works in the domain of Language, and is recorded as handling the task of mathematical reasoning, Quantitative reasoning, Language modeling/generation, Question answering. These are the areas it was designed around; they describe intent rather than a hard boundary.

Deephermes 3 Llama 3 8B Preview— where can I download it?

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

Deephermes 3 Llama 3 8B Preview— 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 1.5 GB. Every figure here assumes the whole model is resident on the card.

Deephermes 3 Llama 3 8B Preview— would two GPUs run it faster?

Capacity adds across cards; throughput does not. The number of cards already holding it on their own: 582. So a second card is rarely the answer here.

Deephermes 3 Llama 3 8B Preview— 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: 4. Bigger cards get the more accurate version, and the quality floor above pins the comparison to one level.

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