DeepSeekMath 7B TPS calculator

Open weights DeepSeek,Tsinghua University,Peking University 7B parameters February 2024

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

509 cards that can run it

818 cards we hold specifications for

Smallest card that fits

Xeon Phi 5110P

8 GB · Q5_K_M · 22.5 tok/s

Fastest card

B200

484 tok/s · 180 GB

Which GPUs can run DeepSeekMath 7B?

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

509 cards match

Calculating

					Needs	Quantisation	Fit
484 tok/s 411–581	B200 NVIDIA	180 GB	8,000 GB/s	Jan 2024	9.6 GB	Q8_0	Comfortable
484 tok/s 411–581	B300 NVIDIA	288 GB	8,000 GB/s	Sep 2025	9.6 GB	Q8_0	Comfortable
387 tok/s 232–618 · low confidence	Radeon Instinct MI350X AMD	288 GB	8,190 GB/s	Jan 2025	9.6 GB	Q8_0	Comfortable
387 tok/s 232–618 · low confidence	Radeon Instinct MI355X AMD	288 GB	8,190 GB/s	Jan 2025	9.6 GB	Q8_0	Comfortable
309 tok/s 185–495 · low confidence	Radeon Instinct MI300 AMD	128 GB	6,550 GB/s	Jan 2023	9.6 GB	Q8_0	Comfortable
296 tok/s 251–355	H200 NVL NVIDIA	141 GB	4,890 GB/s	Nov 2024	9.6 GB	Q8_0	Comfortable
296 tok/s 251–355	H200 SXM 141 GB NVIDIA	141 GB	4,890 GB/s	Nov 2024	9.6 GB	Q8_0	Comfortable
283 tok/s 170–453 · low confidence	Radeon Instinct MI325X AMD	256 GB	6,000 GB/s	Oct 2024	9.6 GB	Q8_0	Comfortable
251 tok/s 151–402 · low confidence	Radeon Instinct MI300A AMD	128 GB	5,325 GB/s	Dec 2023	9.6 GB	Q8_0	Comfortable
251 tok/s 151–402 · low confidence	Radeon Instinct MI300X AMD	192 GB	5,325 GB/s	Dec 2023	9.6 GB	Q8_0	Comfortable
251 tok/s 151–402 · low confidence	Radeon Instinct MI308X AMD	192 GB	5,325 GB/s	Dec 2023	9.6 GB	Q8_0	Comfortable
238 tok/s 203–286	H100 NVL 94 GB NVIDIA	94 GB	3,940 GB/s	Mar 2023	9.6 GB	Q8_0	Comfortable
203 tok/s 173–244	H100 PCIe 96 GB NVIDIA	96 GB	3,360 GB/s	Mar 2023	9.6 GB	Q8_0	Comfortable
203 tok/s 173–244	H100 SXM5 80 GB NVIDIA	80 GB	3,360 GB/s	Oct 2022	9.6 GB	Q8_0	Comfortable
203 tok/s 173–244	H100 SXM5 94 GB NVIDIA	94 GB	3,360 GB/s	Mar 2023	9.6 GB	Q8_0	Comfortable
203 tok/s 173–244	H100 SXM5 96 GB NVIDIA	96 GB	3,360 GB/s	Mar 2023	9.6 GB	Q8_0	Comfortable
203 tok/s 173–244	H800 SXM5 NVIDIA	80 GB	3,360 GB/s	Mar 2023	9.6 GB	Q8_0	Comfortable
161 tok/s 137–193	CMP 170HX 8 GB NVIDIA	8 GB	1,490 GB/s	Sep 2021	7.2 GB	Q5_K_M	Tight
155 tok/s 93–248 · low confidence	Radeon Instinct MI250 AMD	128 GB	3,280 GB/s	Nov 2021	9.6 GB	Q8_0	Comfortable
155 tok/s 93–248 · low confidence	Radeon Instinct MI250X AMD	128 GB	3,280 GB/s	Nov 2021	9.6 GB	Q8_0	Comfortable
137 tok/s 117–165	CMP 170HX 10 GB NVIDIA	10 GB	1,560 GB/s	Sep 2021	8.0 GB	Q6_K	Tight
129 tok/s 77–206 · low confidence	Data Center GPU Max 1550 Intel	128 GB	3,280 GB/s	Jan 2023	9.6 GB	Q8_0	Comfortable
126 tok/s 76–202 · low confidence	Data Center GPU Max Subsystem Intel	128 GB	3,210 GB/s	Jan 2023	9.6 GB	Q8_0	Comfortable
123 tok/s 105–148	A100 SXM4 80 GB NVIDIA	80 GB	2,040 GB/s	Nov 2020	9.6 GB	Q8_0	Comfortable
123 tok/s 105–148	A100X NVIDIA	80 GB	2,040 GB/s	Jun 2021	9.6 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: DeepSeek,Tsinghua University,Peking University
Organisation type: Industry,Academia,Academia
Country: China
Published: 5 February 2024
Authors: Zhihong Shao, Peiyi Wang, Qihao Zhu, Runxin Xu, Junxiao Song, Xiao Bi, Haowei Zhang, Mingchuan Zhang, Y.K. Li, Y. Wu, Daya Guo

What it does

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

Domain: Language
Task: Quantitative reasoning
Base model: DeepSeek Coder 6.7B

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: 7B
Training data: 500,000,000,000 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
Fine-tuning compute: 2.1 × 10²² 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: deepseek-ai

How it is classified

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

Likely above 10²³ FLOP: Yes
Record confidence: Confident

Sources

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

Reference: DeepSeekMath: Pushing the Limits of Mathematical Reasoning in Open Language Models
Last updated: 28 November 2025

The extremes

The ten fastest GPUs that run DeepSeekMath 7B

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 DeepSeekMath 7B

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

Xeon Phi 5110P

Memory needed

7.2 GB

Fastest

484 tok/s

DeepSeekMath 7B is small enough at 7B parameters that hardware is rarely the obstacle — 509 of the cards we track can run it, including cards several years old.

The least hardware that works is a Xeon Phi 5110P. Its 8 GB is enough at Q5_K_M compression, giving roughly 22.5 tokens per second.

A B200 is the fastest we calculate for it: about 484 tokens per second, from 8,000 GB/s of memory bandwidth.

Where it came from

DeepSeekMath 7B was published by DeepSeek,Tsinghua University,Peking University, in China, in February 2024. It comes out of industry,Academia,Academia.

It works in Language, and is recorded as doing quantitative reasoning.

Its starting point was DeepSeek Coder 6.7B — most models at this scale are adapted from an existing base rather than built from nothing.

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. It is published under the deepseek-ai organisation on Hugging Face.

Understanding the speeds

Across every card that can run it, the middle of the range is about 27.2 tokens per second, and 482 of them clear the ten tokens per second that roughly matches reading speed.

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.

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

Training and provenance

Producing it required around 1 × 10²³ FLOP of arithmetic, which is a statement about the training budget rather than about inference.

The training set ran to roughly 500,000,000,000 tokens.

Step by step

How to choose a GPU for DeepSeekMath 7B

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 that can hold DeepSeekMath 7B — around 7.2 GB at Q5_K_M. That figure, not the card's headline performance, is what decides whether it runs.
02

Decide how long your conversations run

The conversation occupies memory too, and grows as it goes. Set the slider to the length you expect: at long context DeepSeekMath 7B can slip off a card that handles short questions easily.
03

Decide how much compression you will accept

Each card runs the least-compressed copy it can hold — Q5_K_M on the smallest card that fits. Setting a floor drops the cards that only manage DeepSeekMath 7B by squeezing it further than you would want.
04

Compare tokens per second, not specifications

Ranking by tokens per second for DeepSeekMath 7B follows memory bandwidth, not core counts, which is why the B200 tops it at 484 tok/s.
05

Look at the headroom, not just the fit

A tight fit runs DeepSeekMath 7B but leaves nothing spare for a longer conversation; comfortable has headroom. If you expect to grow the context, buy for comfortable.
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. Worth a look before buying for DeepSeekMath 7B alone — a card is usually bought for more than one model.

Answers

DeepSeekMath 7B — common questions

Can I run DeepSeekMath 7B on a 16 GB GPU?

Yes. A Tesla V100 SXM2 16 GB with 16 GB runs it at Q8_0, using about 9.6 GB and generating roughly 68.4 tokens per second — a comfortable fit.

Can I run DeepSeekMath 7B on a 24 GB GPU?

Yes. A GeForce RTX 5090 D V2 with 24 GB runs it at Q8_0, using about 9.6 GB and generating roughly 81.1 tokens per second — a comfortable fit.

Is DeepSeekMath 7B open source?

Its weights are published, so DeepSeekMath 7B 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 DeepSeekMath 7B have?

DeepSeekMath 7B has 7B parameters. "Our model is initialized with DeepSeek-Coder-Base-v1.5 7B (Guo et al., 2024) and trained for 500B tokens.". 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 DeepSeekMath 7B?

DeepSeekMath 7B was published by DeepSeek,Tsinghua University,Peking University, based in China, categorised as industry,Academia,Academia.

When was DeepSeekMath 7B released?

DeepSeekMath 7B was published in February 2024. 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 DeepSeekMath 7B used for?

DeepSeekMath 7B works in Language, and is recorded as handling quantitative reasoning. Models frequently carry more than one of each, and the tags describe purpose rather than capability limits.

Where can I download DeepSeekMath 7B?

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

How much compute was used to train DeepSeekMath 7B?

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.

Can I run DeepSeekMath 7B if it does not fit in my GPU?

Only by offloading, which is usually a false economy: the part in system memory drags the whole thing down — the nearest miss we calculate is short by 1.0 GB. Our figures for DeepSeekMath 7B assume it is fully resident.

Would two GPUs run DeepSeekMath 7B faster?

A second card roughly doubles the memory available but not the generation rate. With 509 cards already able to run DeepSeekMath 7B alone, the case for pairing is weak.

Why does the quantisation differ between cards for DeepSeekMath 7B?

A larger card holds a more accurate copy. Across the cards that run DeepSeekMath 7B, 3 compression levels are used; the floor control above pins it to one.

How accurate are these DeepSeekMath 7B speed estimates?

They are calculated from specifications rather than measured, and each carries a range — 411–581 tok/s on the B200, for instance. The same model and card vary by thirty to fifty per cent depending on the inference software and its version.

What GPU do I need to run DeepSeekMath 7B?

The smallest card in our catalogue that holds DeepSeekMath 7B is the Xeon Phi 5110P, with 8 GB of memory. It runs the model at Q5_K_M using about 7.2 GB, and produces roughly 22.5 tokens per second. 509 cards in total can run it.

How fast is DeepSeekMath 7B on a GPU?

It depends on the card. The quickest we calculate is a B200 at about 484 tokens per second; the slowest that still runs it manages considerably less. Reading speed is around ten tokens per second, and 482 of the cards that can run DeepSeekMath 7B clear that.

How much VRAM does DeepSeekMath 7B need?

About 7.2 GB at Q5_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 DeepSeekMath 7B on a 8 GB GPU?

Yes. A CMP 170HX 8 GB with 8 GB runs it at Q5_K_M, using about 7.2 GB and generating roughly 161 tokens per second — a tight fit.

Can I run DeepSeekMath 7B on a 12 GB GPU?

Yes. A GeForce RTX 3080 Ti with 12 GB runs it at Q8_0, using about 9.6 GB and generating roughly 55.2 tokens per second — a tight fit.

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