Qwen3-Omni-30B-A3B TPS calculator

Open weights Alibaba 35.3B parameters September 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 of 818 cards that can run it

Smallest card that fits

RTX A4500

20 GB · IQ4_XS · 105 tok/s

Fastest card

B200

533 tok/s · 180 GB

Which GPUs can run Qwen3-Omni-30B-A3B?

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
533 tok/s 320–853 · low confidence	B200 NVIDIA	180 GB	8,000 GB/s	Jan 2024	36.4 GB	Q8_0	Comfortable
533 tok/s 320–853 · low confidence	B300 NVIDIA	288 GB	8,000 GB/s	Sep 2025	36.4 GB	Q8_0	Comfortable
426 tok/s 255–681 · low confidence	Radeon Instinct MI350X AMD	288 GB	8,190 GB/s	Jan 2025	36.4 GB	Q8_0	Comfortable
426 tok/s 255–681 · low confidence	Radeon Instinct MI355X AMD	288 GB	8,190 GB/s	Jan 2025	36.4 GB	Q8_0	Comfortable
341 tok/s 204–545 · low confidence	Radeon Instinct MI300 AMD	128 GB	6,550 GB/s	Jan 2023	36.4 GB	Q8_0	Comfortable
326 tok/s 196–522 · low confidence	H200 NVL NVIDIA	141 GB	4,890 GB/s	Nov 2024	36.4 GB	Q8_0	Comfortable
326 tok/s 196–522 · low confidence	H200 SXM 141 GB NVIDIA	141 GB	4,890 GB/s	Nov 2024	36.4 GB	Q8_0	Comfortable
312 tok/s 187–499 · low confidence	Radeon Instinct MI325X AMD	256 GB	6,000 GB/s	Oct 2024	36.4 GB	Q8_0	Comfortable
277 tok/s 166–443 · low confidence	Radeon Instinct MI300A AMD	128 GB	5,325 GB/s	Dec 2023	36.4 GB	Q8_0	Comfortable
277 tok/s 166–443 · low confidence	Radeon Instinct MI300X AMD	192 GB	5,325 GB/s	Dec 2023	36.4 GB	Q8_0	Comfortable
277 tok/s 166–443 · low confidence	Radeon Instinct MI308X AMD	192 GB	5,325 GB/s	Dec 2023	36.4 GB	Q8_0	Comfortable
263 tok/s 158–420 · low confidence	H100 NVL 94 GB NVIDIA	94 GB	3,940 GB/s	Mar 2023	36.4 GB	Q8_0	Comfortable
224 tok/s 134–358 · low confidence	H100 PCIe 96 GB NVIDIA	96 GB	3,360 GB/s	Mar 2023	36.4 GB	Q8_0	Comfortable
224 tok/s 134–358 · low confidence	H100 SXM5 80 GB NVIDIA	80 GB	3,360 GB/s	Oct 2022	36.4 GB	Q8_0	Comfortable
224 tok/s 134–358 · low confidence	H100 SXM5 94 GB NVIDIA	94 GB	3,360 GB/s	Mar 2023	36.4 GB	Q8_0	Comfortable
224 tok/s 134–358 · low confidence	H100 SXM5 96 GB NVIDIA	96 GB	3,360 GB/s	Mar 2023	36.4 GB	Q8_0	Comfortable
224 tok/s 134–358 · low confidence	H800 SXM5 NVIDIA	80 GB	3,360 GB/s	Mar 2023	36.4 GB	Q8_0	Comfortable
206 tok/s 124–330 · low confidence	GeForce RTX 5090 D V2 NVIDIA	24 GB	1,340 GB/s	Aug 2025	20.0 GB	Q4_K_M	Tight
188 tok/s 113–300 · low confidence	A30X NVIDIA	24 GB	1,220 GB/s	Apr 2021	20.0 GB	Q4_K_M	Tight
181 tok/s 109–290 · low confidence	DRIVE A100 PROD NVIDIA	32 GB	1,870 GB/s	May 2020	28.2 GB	Q6_K	Tight
181 tok/s 109–290 · low confidence	GRID A100A NVIDIA	32 GB	1,870 GB/s	May 2020	28.2 GB	Q6_K	Tight
173 tok/s 104–277 · low confidence	GeForce RTX 5090 NVIDIA	32 GB	1,790 GB/s	Jan 2025	28.2 GB	Q6_K	Tight
173 tok/s 104–277 · low confidence	GeForce RTX 5090 D NVIDIA	32 GB	1,790 GB/s	Jan 2025	28.2 GB	Q6_K	Tight
171 tok/s 102–273 · low confidence	Radeon Instinct MI250 AMD	128 GB	3,280 GB/s	Nov 2021	36.4 GB	Q8_0	Comfortable
171 tok/s 102–273 · low confidence	Radeon Instinct MI250X AMD	128 GB	3,280 GB/s	Nov 2021	36.4 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: 22 September 2025
Authors: Jin Xu, Zhifang Guo, Hangrui Hu, Yunfei Chu, Xiong Wang, Jinzheng He, Yuxuan Wang, Xian Shi, Ting He, Xinfa Zhu, Yuanjun Lv, Yongqi Wang, Dake Guo, He Wang, Linhan Ma, Pei Zhang, Xinyu Zhang, Hongkun Hao, Zishan Guo, Baosong Yang, Bin Zhang, Ziyang Ma, Xipin Wei, Shuai Bai, Keqin Chen, Xuejing Liu, Peng Wang, Mingkun Yang, Dayiheng Liu, Xingzhang Ren, Bo Zheng, Rui Men, Fan Zhou, Bowen Yu, Jianxin…

What it does

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

Domain: Multimodal, Language, Vision, Speech, Video
Task: Language modeling/generation, Question answering, Visual question answering, Image captioning, Video description, Speech recognition (ASR), Speech synthesis, Speech-to-text, Text-to-speech (TTS)

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: 35.3B
Training data: 2,000,000,000,000 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: 3.6 × 10²² 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.

Why it is tracked: SOTA improvement
Record confidence: Confident

Sources

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

Reference: Qwen3-Omni Technical Report
Last updated: 18 December 2025

The extremes

The ten fastest GPUs for Qwen3-Omni-30B-A3B

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-Omni-30B-A3B

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

RTX A4500

Memory needed

17.9 GB

Fastest

533 tok/s

With 35.3B parameters, Qwen3-Omni-30B-A3B 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 smallest card that holds it is the RTX A4500 with 20 GB, running it at IQ4_XS and producing around 105 tokens per second.

Top of the range is the B200, at roughly 533 tokens per second thanks to 8,000 GB/s of bandwidth.

Where it came from

Qwen3-Omni-30B-A3B was published by Alibaba, in China, in September 2025. The organisation is categorised as industry.

It works in Multimodal, Language, Vision, Speech, Video, and is recorded as doing language modeling/generation, Question answering, Visual question answering, Image captioning, Video description, Speech recognition (ASR), Speech synthesis, Speech-to-text, Text-to-speech (TTS).

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 Qwen organisation on Hugging Face.

Understanding the speeds

Across every card that can run it, the middle of the range is about 104.0 tokens per second, and 132 of them clear the ten tokens per second that roughly matches reading speed.

This is a mixture-of-experts model, which routes each token through only part of itself. It therefore generates far faster than its total size suggests — while still needing every parameter resident in memory, so it is quick without being cheap to hold.

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.

How it was trained

The training run consumed about 3.6 × 10²² FLOP. That figure describes the cost of creating it and has no bearing on how quickly it generates text.

The training set ran to roughly 2,000,000,000,000 tokens.

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

Step by step

How to choose a GPU for Qwen3-Omni-30B-A3B

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

Look at what Qwen3-Omni-30B-A3B actually needs — around 17.9 GB at IQ4_XS. No amount of processing power compensates for a card that cannot hold it.
02

Match the context to your actual use

Set the context to what you will actually use. The cache grows with the conversation, and it is the usual reason Qwen3-Omni-30B-A3B stops fitting a card that seemed fine.
03

Choose how far you will compress it

Each card runs the least-compressed copy it can hold — IQ4_XS on the smallest card that fits. Setting a floor drops the cards that only manage Qwen3-Omni-30B-A3B by squeezing it further than you would want.
04

Rank by throughput rather than spec sheet

The speed ordering for Qwen3-Omni-30B-A3B is effectively an ordering by memory bandwidth, which is why the B200 tops it at 533 tok/s.
05

Look at the headroom, not just the fit

The fit column separates cards that just manage Qwen3-Omni-30B-A3B from those with room to spare. Buy for the second if the context might grow.
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 Qwen3-Omni-30B-A3B alone — a card is usually bought for more than one model.

Answers

Qwen3-Omni-30B-A3B — common questions

Why does the quantisation differ between cards for Qwen3-Omni-30B-A3B?

A larger card holds a more accurate copy. Across the cards that run Qwen3-Omni-30B-A3B, 5 compression levels are used; the floor control above pins it to one.

How accurate are these Qwen3-Omni-30B-A3B speed estimates?

Every figure is derived from memory bandwidth and model size, not benchmarked. That is why each is published as a range such as 320–853 tok/s on the B200 rather than a single number.

What GPU do I need to run Qwen3-Omni-30B-A3B?

The smallest card in our catalogue that holds Qwen3-Omni-30B-A3B is the RTX A4500, with 20 GB of memory. It runs the model at IQ4_XS using about 17.9 GB, and produces roughly 105 tokens per second. 132 cards in total can run it.

How fast is Qwen3-Omni-30B-A3B on a GPU?

It depends on the card. The quickest we calculate is a B200 at about 533 tokens per second; the slowest that still runs it manages considerably less. Reading speed is around ten tokens per second, and 132 of the cards that can run Qwen3-Omni-30B-A3B clear that.

How much VRAM does Qwen3-Omni-30B-A3B need?

About 17.9 GB at IQ4_XS 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 Qwen3-Omni-30B-A3B on a 24 GB GPU?

Yes. A GeForce RTX 5090 D V2 with 24 GB runs it at Q4_K_M, using about 20.0 GB and generating roughly 206 tokens per second — a tight fit.

Is Qwen3-Omni-30B-A3B open source?

Its weights are published, so Qwen3-Omni-30B-A3B 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 Qwen3-Omni-30B-A3B have?

Qwen3-Omni-30B-A3B has 35.3B parameters. Audio Encoder AuT 650M Vision Encoder SigLIP2-So400M 540M Thinker MoE Transformer 30B-A3B Talker MoE Transformer 3B-A0.3B MTP Dense Transformer 80M Code2wav ConvNet 200M. 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 Qwen3-Omni-30B-A3B?

Qwen3-Omni-30B-A3B was published by Alibaba, based in China, categorised as industry.

When was Qwen3-Omni-30B-A3B released?

Qwen3-Omni-30B-A3B was published in September 2025.

What is Qwen3-Omni-30B-A3B used for?

Qwen3-Omni-30B-A3B works in Multimodal, Language, Vision, Speech, Video, and is recorded as handling language modeling/generation, Question answering, Visual question answering, Image captioning, Video description, Speech recognition (ASR), Speech synthesis, Speech-to-text, Text-to-speech (TTS). 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.

Where can I download Qwen3-Omni-30B-A3B?

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 Qwen3-Omni-30B-A3B?

Around 3.6 × 10²² 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 Qwen3-Omni-30B-A3B if it does not fit in my GPU?

It can be split between the card and system memory, but Qwen3-Omni-30B-A3B generates painfully slowly that way — the nearest miss we calculate is short by 5.6 GB. Nothing on this page assumes offloading.

Would two GPUs run Qwen3-Omni-30B-A3B faster?

A second card roughly doubles the memory available but not the generation rate. With 132 cards already able to run Qwen3-Omni-30B-A3B alone, the case for pairing is weak.

Source

Original publication

Record last updated 18 December 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