Qwen3-235B-A22B-Thinking (Jul 2025) TPS calculator

Open weights Alibaba 235B parameters July 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

17 cards that can run it

818 cards we hold specifications for

Smallest card that fits

Radeon Instinct MI250

128 GB · Q3_K_M · 69.1 tok/s

Fastest card

B200

143 tok/s · 180 GB

Which GPUs can run Qwen3-235B-A22B-Thinking (Jul 2025)?

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

17 cards match

Calculating

					Needs	Quantisation	Fit
143 tok/s 86–229 · low confidence	B200 NVIDIA	180 GB	8,000 GB/s	Jan 2024	157.1 GB	Q5_K_M	Tight
138 tok/s 83–221 · low confidence	Radeon Instinct MI300 AMD	128 GB	6,550 GB/s	Jan 2023	102.4 GB	Q3_K_M	Tight
120 tok/s 72–192 · low confidence	H200 NVL NVIDIA	141 GB	4,890 GB/s	Nov 2024	116.0 GB	IQ4_XS	Tight
120 tok/s 72–192 · low confidence	H200 SXM 141 GB NVIDIA	141 GB	4,890 GB/s	Nov 2024	116.0 GB	IQ4_XS	Tight
112 tok/s 67–180 · low confidence	Radeon Instinct MI300A AMD	128 GB	5,325 GB/s	Dec 2023	102.4 GB	Q3_K_M	Tight
80.1 tok/s 48–128 · low confidence	B300 NVIDIA	288 GB	8,000 GB/s	Sep 2025	239.1 GB	Q8_0	Tight
74.3 tok/s 45–119 · low confidence	Radeon Instinct MI300X AMD	192 GB	5,325 GB/s	Dec 2023	157.1 GB	Q5_K_M	Tight
74.3 tok/s 45–119 · low confidence	Radeon Instinct MI308X AMD	192 GB	5,325 GB/s	Dec 2023	157.1 GB	Q5_K_M	Tight
69.1 tok/s 41–111 · low confidence	Radeon Instinct MI250 AMD	128 GB	3,280 GB/s	Nov 2021	102.4 GB	Q3_K_M	Tight
69.1 tok/s 41–111 · low confidence	Radeon Instinct MI250X AMD	128 GB	3,280 GB/s	Nov 2021	102.4 GB	Q3_K_M	Tight
68.1 tok/s 41–109 · low confidence	Radeon Instinct MI325X AMD	256 GB	6,000 GB/s	Oct 2024	184.4 GB	Q6_K	Comfortable
64.0 tok/s 38–102 · low confidence	Radeon Instinct MI350X AMD	288 GB	8,190 GB/s	Jan 2025	239.1 GB	Q8_0	Tight
64.0 tok/s 38–102 · low confidence	Radeon Instinct MI355X AMD	288 GB	8,190 GB/s	Jan 2025	239.1 GB	Q8_0	Tight
57.6 tok/s 35–92 · low confidence	Data Center GPU Max 1550 Intel	128 GB	3,280 GB/s	Jan 2023	102.4 GB	Q3_K_M	Tight
56.4 tok/s 34–90 · low confidence	Data Center GPU Max Subsystem Intel	128 GB	3,210 GB/s	Jan 2023	102.4 GB	Q3_K_M	Tight
7.4 tok/s 4–12 · low confidence	GB10 NVIDIA	128 GB	273 GB/s	Oct 2025	102.4 GB	Q3_K_M	Tight
7.4 tok/s 4–12 · low confidence	Jetson T5000 NVIDIA	128 GB	273 GB/s	Aug 2025	102.4 GB	Q3_K_M	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: Alibaba
Organisation type: Industry
Country: China
Published: 25 July 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, Quantitative reasoning, Code generation, Translation

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: 235B
Training data: 36,000,000,000,000 tokens
Epochs: 1

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: 4.8 × 10²⁴ 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: 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: Discretionary
Record confidence: Likely

Sources

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

Reference: Qwen3: Think Deeper, Act Faster
Last updated: 12 February 2026

The extremes

The ten fastest GPUs that run Qwen3-235B-A22B-Thinking (Jul 2025)

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 Qwen3-235B-A22B-Thinking (Jul 2025)

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

Radeon Instinct MI250

Memory needed

102.4 GB

Fastest

143 tok/s

Qwen3-235B-A22B-Thinking (Jul 2025) reaches a parameter count of 235B. That is beyond what any single graphics card holds. Running it means either splitting it across several cards or renting hardware built for the job, and every card able to hold it alone is a datacentre part. The number that can: 17.

The smallest card that holds it is Radeon Instinct MI250, with a memory capacity of 128 GB, running it at a compression of Q3_K_M and producing around 69.1 tokens per second.

At the other end sits B200, generating roughly 143 tokens per second on the strength of a memory bandwidth of 8,000 GB/s.

Where it came from

Qwen3-235B-A22B-Thinking (Jul 2025) was published by Alibaba, in the country recorded as China, during July 2025. The publishing organisation is categorised as industry.

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

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

Understanding the speeds

Half the cards that hold it manage more than 69.1 tokens per second. Producing text faster than most people read it: 15 of them.

Because it routes each token through a subset of its weights, it produces text at the pace of a much smaller model. The catch is memory: all of it still has to fit, so the speed is a bonus rather than a discount on hardware.

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 4.8 × 10²⁴ FLOP. That figure measures what producing the model cost, and has no bearing on how fast it answers.

Training consumed a corpus of around 36,000,000,000,000 tokens of text.

It is tracked in the underlying dataset for one reason in particular: discretionary.

Step by step

How to choose a GPU for Qwen3-235B-A22B-Thinking (Jul 2025)

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 able to hold Qwen3-235B-A22B-Thinking (Jul 2025), needing around 102.4 GB at a compression of Q3_K_M. Capacity is the gate — a card either holds it or it does not.
02

Match the context to your actual use

The conversation occupies memory too, and grows as it goes. Set the slider to the length you expect, because at long context a card that handles short questions easily can be dropped by Qwen3-235B-A22B-Thinking (Jul 2025).
03

Set a quality floor

The quantisation column varies by card, because a bigger card holds a more accurate copy, 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

Sort by speed

Ranking by tokens per second follows memory bandwidth rather than core counts, for Qwen3-235B-A22B-Thinking (Jul 2025). It will not match a gaming ordering, because generation is bound by memory bandwidth. The card topping the list is B200, at 143 tok/s.
05

Look at the headroom, not just the fit

A tight fit runs, but leaves nothing spare for a longer conversation, in the case of Qwen3-235B-A22B-Thinking (Jul 2025). 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

Check the card from the other side

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 Qwen3-235B-A22B-Thinking (Jul 2025).

Answers

Qwen3-235B-A22B-Thinking (Jul 2025) — common questions

Qwen3-235B-A22B-Thinking (Jul 2025)— would two GPUs run it faster?

Capacity adds across cards; throughput does not. The number of cards already holding it on their own: 17. So a second card is rarely the answer here.

Qwen3-235B-A22B-Thinking (Jul 2025)— 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.

Qwen3-235B-A22B-Thinking (Jul 2025)— how accurate are these speed estimates?

Every figure is derived from memory bandwidth and model size, not benchmarked, which is why each is published as a range rather than a single number. One example: 86–229 tok/s on B200. The same model and card vary by thirty to fifty per cent depending on the inference software and its version.

Qwen3-235B-A22B-Thinking (Jul 2025)— what GPU do I need to run it?

The smallest card in our catalogue that holds it is Radeon Instinct MI250, with a memory capacity of 128 GB. It runs the model at a compression of Q3_K_M using about 102.4 GB, and produces roughly 69.1 tokens per second. The number of cards able to run it in total: 17.

Qwen3-235B-A22B-Thinking (Jul 2025)— how fast is it on a GPU?

It depends on the card. The quickest we calculate is B200, at about 143 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: 15.

Qwen3-235B-A22B-Thinking (Jul 2025)— how much VRAM does it need?

It needs about 102.4 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.

Qwen3-235B-A22B-Thinking (Jul 2025)— 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.

Qwen3-235B-A22B-Thinking (Jul 2025)— how many parameters does it have?

It has a parameter count of 235B. 235 billion total parameters and 22 billion activated parameters Number of Layers: 94 Number of Attention Heads (GQA): 64 for Q and 4 for KV Number of Experts: 128 Number of Activated Experts: 8 Context Length: 32,768 natively and 131,072 tokens with YaRN. 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.

Qwen3-235B-A22B-Thinking (Jul 2025)— who created it?

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

Qwen3-235B-A22B-Thinking (Jul 2025)— when was it released?

It was published in July 2025.

Qwen3-235B-A22B-Thinking (Jul 2025)— 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, Quantitative reasoning, Code generation, Translation. A model can carry several of each, so these are the areas it was built for rather than a limit on what it will attempt.

Qwen3-235B-A22B-Thinking (Jul 2025)— where can I download it?

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

Qwen3-235B-A22B-Thinking (Jul 2025)— how much compute was used to train it?

Training consumed around 4.8 × 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.

Qwen3-235B-A22B-Thinking (Jul 2025)— 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 43.3 GB. Every figure here assumes the whole model is resident on the card.

Source

Original publication

Record last updated 12 February 2026

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