QwQ-32B TPS calculator

Open weights Alibaba 32.5B parameters March 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

132 cards that can run it

818 cards we hold specifications for

Smallest card that fits

RTX A4500

20 GB · Q3_K_M · 22.5 tok/s

Fastest card

B200

104 tok/s · 180 GB

Which GPUs can run QwQ-32B?

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

132 cards match

Calculating

					Needs	Quantisation	Fit
104 tok/s 63–167 · low confidence	B200 NVIDIA	180 GB	8,000 GB/s	Jan 2024	35.5 GB	Q8_0	Comfortable
104 tok/s 63–167 · low confidence	B300 NVIDIA	288 GB	8,000 GB/s	Sep 2025	35.5 GB	Q8_0	Comfortable
83.3 tok/s 50–133 · low confidence	Radeon Instinct MI350X AMD	288 GB	8,190 GB/s	Jan 2025	35.5 GB	Q8_0	Comfortable
83.3 tok/s 50–133 · low confidence	Radeon Instinct MI355X AMD	288 GB	8,190 GB/s	Jan 2025	35.5 GB	Q8_0	Comfortable
66.6 tok/s 40–107 · low confidence	Radeon Instinct MI300 AMD	128 GB	6,550 GB/s	Jan 2023	35.5 GB	Q8_0	Comfortable
63.7 tok/s 38–102 · low confidence	H200 NVL NVIDIA	141 GB	4,890 GB/s	Nov 2024	35.5 GB	Q8_0	Comfortable
63.7 tok/s 38–102 · low confidence	H200 SXM 141 GB NVIDIA	141 GB	4,890 GB/s	Nov 2024	35.5 GB	Q8_0	Comfortable
61.0 tok/s 37–98 · low confidence	Radeon Instinct MI325X AMD	256 GB	6,000 GB/s	Oct 2024	35.5 GB	Q8_0	Comfortable
54.1 tok/s 32–87 · low confidence	Radeon Instinct MI300A AMD	128 GB	5,325 GB/s	Dec 2023	35.5 GB	Q8_0	Comfortable
54.1 tok/s 32–87 · low confidence	Radeon Instinct MI300X AMD	192 GB	5,325 GB/s	Dec 2023	35.5 GB	Q8_0	Comfortable
54.1 tok/s 32–87 · low confidence	Radeon Instinct MI308X AMD	192 GB	5,325 GB/s	Dec 2023	35.5 GB	Q8_0	Comfortable
51.3 tok/s 31–82 · low confidence	H100 NVL 94 GB NVIDIA	94 GB	3,940 GB/s	Mar 2023	35.5 GB	Q8_0	Comfortable
43.8 tok/s 26–70 · low confidence	H100 PCIe 96 GB NVIDIA	96 GB	3,360 GB/s	Mar 2023	35.5 GB	Q8_0	Comfortable
43.8 tok/s 26–70 · low confidence	H100 SXM5 80 GB NVIDIA	80 GB	3,360 GB/s	Oct 2022	35.5 GB	Q8_0	Comfortable
43.8 tok/s 26–70 · low confidence	H100 SXM5 94 GB NVIDIA	94 GB	3,360 GB/s	Mar 2023	35.5 GB	Q8_0	Comfortable
43.8 tok/s 26–70 · low confidence	H100 SXM5 96 GB NVIDIA	96 GB	3,360 GB/s	Mar 2023	35.5 GB	Q8_0	Comfortable
43.8 tok/s 26–70 · low confidence	H800 SXM5 NVIDIA	80 GB	3,360 GB/s	Mar 2023	35.5 GB	Q8_0	Comfortable
40.3 tok/s 24–65 · low confidence	GeForce RTX 5090 D V2 NVIDIA	24 GB	1,340 GB/s	Aug 2025	20.4 GB	Q4_K_M	Tight
36.7 tok/s 22–59 · low confidence	A30X NVIDIA	24 GB	1,220 GB/s	Apr 2021	20.4 GB	Q4_K_M	Tight
35.4 tok/s 21–57 · low confidence	DRIVE A100 PROD NVIDIA	32 GB	1,870 GB/s	May 2020	27.9 GB	Q6_K	Tight
35.4 tok/s 21–57 · low confidence	GRID A100A NVIDIA	32 GB	1,870 GB/s	May 2020	27.9 GB	Q6_K	Tight
33.9 tok/s 20–54 · low confidence	GeForce RTX 5090 NVIDIA	32 GB	1,790 GB/s	Jan 2025	27.9 GB	Q6_K	Tight
33.9 tok/s 20–54 · low confidence	GeForce RTX 5090 D NVIDIA	32 GB	1,790 GB/s	Jan 2025	27.9 GB	Q6_K	Tight
33.3 tok/s 20–53 · low confidence	Radeon Instinct MI250 AMD	128 GB	3,280 GB/s	Nov 2021	35.5 GB	Q8_0	Comfortable
33.3 tok/s 20–53 · low confidence	Radeon Instinct MI250X AMD	128 GB	3,280 GB/s	Nov 2021	35.5 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: Alibaba
Organisation type: Industry
Country: China
Published: 6 March 2025
Authors: Qwen Team

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, Quantitative reasoning, Code generation
Base model: Qwen2.5-Coder (32B)
Numerical format: BF16

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: 32.5B
Training data: 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: 3.5 × 10²⁴ 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 (unrestricted)
Training code: Unreleased
Hugging Face: Qwen

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
Why it is tracked: SOTA improvement
Record confidence: Speculative

Sources

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

Reference: QwQ-32B: Embracing the Power of Reinforcement Learning
Last updated: 28 November 2025

The extremes

The ten fastest GPUs that run QwQ-32B

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 QwQ-32B

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

RTX A4500

Memory needed

16.6 GB

Fastest

104 tok/s

With 32.5B parameters, QwQ-32B lands in the range a serious desktop card can handle once the weights are compressed. 132 of the cards we track can run it.

The least hardware that works is a RTX A4500. Its 20 GB is enough at Q3_K_M compression, giving roughly 22.5 tokens per second.

At the other end, a B200 generates roughly 104 tokens per second on it, on the strength of 8,000 GB/s of memory bandwidth.

Background

QwQ-32B was published by Alibaba, in China, in March 2025. The organisation is categorised as industry.

It works in Language, and is recorded as doing language modeling/generation, Question answering, Quantitative reasoning, Code generation.

It builds on Qwen2.5-Coder (32B), which is why it shares that model's general shape and size.

The weights are published, so it can be downloaded and run on your own hardware indefinitely, offline, with no account attached. It is published under the Qwen organisation on Hugging Face.

Reading the throughput figures

The median result is around 20.3 tokens per second; 102 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.

Without the attention layout on record, the memory column is an approximation. It is close enough to choose hardware by, and least reliable at long context.

Training and provenance

The training run consumed about 3.5 × 10²⁴ FLOP. That figure describes the cost of creating it and has no bearing on how quickly it generates text.

The reason it appears in this catalogue at all is sOTA improvement.

Step by step

How to choose a GPU for QwQ-32B

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 QwQ-32B actually needs — around 16.6 GB at Q3_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 QwQ-32B.
03

Set a quality floor

Compression is what makes QwQ-32B fit smaller cards, at some cost in accuracy — Q3_K_M on the smallest card that fits. A minimum quality removes the ones that go too far.
04

Sort by speed

Sort by speed to see how cards rank for QwQ-32B. It will not match a gaming ordering — generation is bound by memory bandwidth, which is why the B200 tops it at 104 tok/s.
05

Read the fit column last

A tight fit runs QwQ-32B 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

Each card page repeats this sweep for every model we hold. It answers what else the hardware is good for, beyond QwQ-32B.

Answers

QwQ-32B — common questions

Would two GPUs run QwQ-32B faster?

Capacity adds across cards; throughput does not. Since 132 of the cards we track already hold QwQ-32B on their own, a second card is rarely the answer here.

Why does the quantisation differ between cards for QwQ-32B?

A larger card holds a more accurate copy. Across the cards that run QwQ-32B, 5 compression levels are used; the floor control above pins it to one.

How accurate are these QwQ-32B speed estimates?

These are estimates with real error bars. The fastest result here, 63–167 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 QwQ-32B?

The smallest card in our catalogue that holds QwQ-32B is the RTX A4500, with 20 GB of memory. It runs the model at Q3_K_M using about 16.6 GB, and produces roughly 22.5 tokens per second. 132 cards in total can run it.

How fast is QwQ-32B on a GPU?

It depends on the card. The quickest we calculate is a B200 at about 104 tokens per second; the slowest that still runs it manages considerably less. Reading speed is around ten tokens per second, and 102 of the cards that can run QwQ-32B clear that.

How much VRAM does QwQ-32B need?

About 16.6 GB at Q3_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 QwQ-32B on a 24 GB GPU?

Yes. A GeForce RTX 5090 D V2 with 24 GB runs it at Q4_K_M, using about 20.4 GB and generating roughly 40.3 tokens per second — a tight fit.

Is QwQ-32B open source?

Its weights are published, so QwQ-32B 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 QwQ-32B have?

QwQ-32B has 32.5B parameters. Architecture: transformers with RoPE, SwiGLU, RMSNorm, and Attention QKV bias Number of Parameters: 32.5B Number of Paramaters (Non-Embedding): 31.0B Number of Layers: 64 Number of Attention Heads (GQA): 40 for Q and 8 for KV. 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 QwQ-32B?

QwQ-32B was published by Alibaba, based in China, categorised as industry.

When was QwQ-32B released?

QwQ-32B was published in March 2025.

What is QwQ-32B used for?

QwQ-32B works in Language, and is recorded as handling language modeling/generation, Question answering, Quantitative reasoning, Code generation. Models frequently carry more than one of each, and the tags describe purpose rather than capability limits.

Where can I download QwQ-32B?

Its weights are published under the Qwen 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 QwQ-32B?

Around 3.5 × 10²⁴ 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 QwQ-32B if it does not fit in my GPU?

It can be split between the card and system memory, but QwQ-32B generates painfully slowly that way — the nearest miss we calculate is short by 6.0 GB. Nothing on this page assumes offloading.

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