Ovis1.6-Gemma2-9B TPS calculator

Open weights Alibaba 10.2B parameters September 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 · Q4_K_M · 19.9 tok/s

Fastest card

B200

332 tok/s · 180 GB

Which GPUs can run Ovis1.6-Gemma2-9B?

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
332 tok/s 199–531 · low confidence	B200 NVIDIA	180 GB	8,000 GB/s	Jan 2024	11.6 GB	Q8_0	Comfortable
332 tok/s 199–531 · low confidence	B300 NVIDIA	288 GB	8,000 GB/s	Sep 2025	11.6 GB	Q8_0	Comfortable
265 tok/s 159–424 · low confidence	Radeon Instinct MI350X AMD	288 GB	8,190 GB/s	Jan 2025	11.6 GB	Q8_0	Comfortable
265 tok/s 159–424 · low confidence	Radeon Instinct MI355X AMD	288 GB	8,190 GB/s	Jan 2025	11.6 GB	Q8_0	Comfortable
212 tok/s 127–339 · low confidence	Radeon Instinct MI300 AMD	128 GB	6,550 GB/s	Jan 2023	11.6 GB	Q8_0	Comfortable
203 tok/s 122–325 · low confidence	H200 NVL NVIDIA	141 GB	4,890 GB/s	Nov 2024	11.6 GB	Q8_0	Comfortable
203 tok/s 122–325 · low confidence	H200 SXM 141 GB NVIDIA	141 GB	4,890 GB/s	Nov 2024	11.6 GB	Q8_0	Comfortable
194 tok/s 117–311 · low confidence	Radeon Instinct MI325X AMD	256 GB	6,000 GB/s	Oct 2024	11.6 GB	Q8_0	Comfortable
172 tok/s 103–276 · low confidence	Radeon Instinct MI300A AMD	128 GB	5,325 GB/s	Dec 2023	11.6 GB	Q8_0	Comfortable
172 tok/s 103–276 · low confidence	Radeon Instinct MI300X AMD	192 GB	5,325 GB/s	Dec 2023	11.6 GB	Q8_0	Comfortable
172 tok/s 103–276 · low confidence	Radeon Instinct MI308X AMD	192 GB	5,325 GB/s	Dec 2023	11.6 GB	Q8_0	Comfortable
164 tok/s 98–262 · low confidence	H100 NVL 94 GB NVIDIA	94 GB	3,940 GB/s	Mar 2023	11.6 GB	Q8_0	Comfortable
143 tok/s 86–229 · low confidence	CMP 170HX 8 GB NVIDIA	8 GB	1,490 GB/s	Sep 2021	6.9 GB	Q4_K_M	Tight
140 tok/s 84–223 · low confidence	H100 PCIe 96 GB NVIDIA	96 GB	3,360 GB/s	Mar 2023	11.6 GB	Q8_0	Comfortable
140 tok/s 84–223 · low confidence	H100 SXM5 80 GB NVIDIA	80 GB	3,360 GB/s	Oct 2022	11.6 GB	Q8_0	Comfortable
140 tok/s 84–223 · low confidence	H100 SXM5 94 GB NVIDIA	94 GB	3,360 GB/s	Mar 2023	11.6 GB	Q8_0	Comfortable
140 tok/s 84–223 · low confidence	H100 SXM5 96 GB NVIDIA	96 GB	3,360 GB/s	Mar 2023	11.6 GB	Q8_0	Comfortable
140 tok/s 84–223 · low confidence	H800 SXM5 NVIDIA	80 GB	3,360 GB/s	Mar 2023	11.6 GB	Q8_0	Comfortable
116 tok/s 69–185 · low confidence	CMP 170HX 10 GB NVIDIA	10 GB	1,560 GB/s	Sep 2021	8.1 GB	Q5_K_M	Tight
106 tok/s 64–170 · low confidence	Radeon Instinct MI250 AMD	128 GB	3,280 GB/s	Nov 2021	11.6 GB	Q8_0	Comfortable
106 tok/s 64–170 · low confidence	Radeon Instinct MI250X AMD	128 GB	3,280 GB/s	Nov 2021	11.6 GB	Q8_0	Comfortable
88.5 tok/s 53–142 · low confidence	Data Center GPU Max 1550 Intel	128 GB	3,280 GB/s	Jan 2023	11.6 GB	Q8_0	Comfortable
86.6 tok/s 52–139 · low confidence	Data Center GPU Max Subsystem Intel	128 GB	3,210 GB/s	Jan 2023	11.6 GB	Q8_0	Comfortable
84.7 tok/s 51–136 · low confidence	A100 SXM4 80 GB NVIDIA	80 GB	2,040 GB/s	Nov 2020	11.6 GB	Q8_0	Comfortable
84.7 tok/s 51–136 · low confidence	A100X NVIDIA	80 GB	2,040 GB/s	Jun 2021	11.6 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: Alibaba
Organisation type: Industry
Country: China
Published: 16 September 2024

What it does

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

Domain: Multimodal, Language, Vision
Task: Language modeling/generation, Visual question answering
Base model: Gemma 2 9B,SigLIP 400M

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: 10.2B
Training data: tokens

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)
Hugging Face: AIDC-AI

How it is classified

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

Record confidence: Confident

Sources

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

Reference: Ovis1.6-Gemma2-9B
Last updated: 28 November 2025

The extremes

The ten fastest GPUs that run Ovis1.6-Gemma2-9B

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 Ovis1.6-Gemma2-9B

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

Xeon Phi 5110P

Memory needed

6.9 GB

Fastest

332 tok/s

Ovis1.6-Gemma2-9B is small enough at 10.2B parameters that hardware is rarely the obstacle — 509 of the cards we track can run it, including cards several years old.

At the low end, a Xeon Phi 5110P handles it — 8 GB, at Q4_K_M, for about 19.9 tokens per second.

The quickest result comes from a B200 at around 332 tokens per second — its 8,000 GB/s of bandwidth is what buys that.

Background

Ovis1.6-Gemma2-9B was published by Alibaba, in China, in September 2024. The organisation is categorised as industry.

It works in Multimodal, Language, Vision, and is recorded as doing language modeling/generation, Visual question answering.

It is derived from Gemma 2 9B,SigLIP 400M rather than trained from scratch, which is the usual way a specialised model is produced.

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 AIDC-AI organisation on Hugging Face.

Reading the throughput figures

Across every card that can run it, the middle of the range is about 21.5 tokens per second, and 467 of them clear the ten tokens per second that roughly matches reading speed.

It is a dense model, so every parameter is read for every token produced. That makes speed track memory bandwidth almost exactly — a card with twice the bandwidth generates roughly twice as fast.

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.

Step by step

How to choose a GPU for Ovis1.6-Gemma2-9B

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 Ovis1.6-Gemma2-9B — around 6.9 GB at Q4_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 Ovis1.6-Gemma2-9B can slip off a card that handles short questions easily.
03

Choose how far you will compress it

Compression is what makes Ovis1.6-Gemma2-9B fit smaller cards, at some cost in accuracy — Q4_K_M on the smallest card that fits. A minimum quality removes the ones that go too far.
04

Sort by speed

Ranking by tokens per second for Ovis1.6-Gemma2-9B follows memory bandwidth, not core counts, which is why the B200 tops it at 332 tok/s.
05

Look at the headroom, not just the fit

The fit column separates cards that just manage Ovis1.6-Gemma2-9B from those with room to spare. Buy for the second if the context might grow.
06

Check the card from the other side

Every card name links to its own page, which runs the same calculation across the whole model catalogue. Worth a look before buying for Ovis1.6-Gemma2-9B alone — a card is usually bought for more than one model.

Answers

Ovis1.6-Gemma2-9B — common questions

Can I run Ovis1.6-Gemma2-9B if it does not fit in my GPU?

It can be split between the card and system memory, but Ovis1.6-Gemma2-9B generates painfully slowly that way — the nearest miss we calculate is short by 1.5 GB. Nothing on this page assumes offloading.

Would two GPUs run Ovis1.6-Gemma2-9B faster?

Two cards buy memory rather than speed. That matters for Ovis1.6-Gemma2-9B only if one card cannot hold it — 509 can, so a second adds little.

Why does the quantisation differ between cards for Ovis1.6-Gemma2-9B?

Each card is shown running the least-compressed copy it can hold, and Ovis1.6-Gemma2-9B appears at 4 different compression levels across the cards that fit it. Bigger cards get the more accurate version.

How accurate are these Ovis1.6-Gemma2-9B speed estimates?

These are estimates with real error bars. The fastest result here, 199–531 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 Ovis1.6-Gemma2-9B?

The smallest card in our catalogue that holds Ovis1.6-Gemma2-9B is the Xeon Phi 5110P, with 8 GB of memory. It runs the model at Q4_K_M using about 6.9 GB, and produces roughly 19.9 tokens per second. 509 cards in total can run it.

How fast is Ovis1.6-Gemma2-9B on a GPU?

It depends on the card. The quickest we calculate is a B200 at about 332 tokens per second; the slowest that still runs it manages considerably less. Reading speed is around ten tokens per second, and 467 of the cards that can run Ovis1.6-Gemma2-9B clear that.

How much VRAM does Ovis1.6-Gemma2-9B need?

About 6.9 GB at Q4_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 Ovis1.6-Gemma2-9B on a 8 GB GPU?

Yes. A CMP 170HX 8 GB with 8 GB runs it at Q4_K_M, using about 6.9 GB and generating roughly 143 tokens per second — a tight fit.

Can I run Ovis1.6-Gemma2-9B on a 12 GB GPU?

Yes. A GeForce RTX 3080 Ti with 12 GB runs it at Q6_K, using about 9.2 GB and generating roughly 55.1 tokens per second — a tight fit.

Can I run Ovis1.6-Gemma2-9B on a 16 GB GPU?

Yes. A Tesla V100 SXM2 16 GB with 16 GB runs it at Q8_0, using about 11.6 GB and generating roughly 46.9 tokens per second — a comfortable fit.

Can I run Ovis1.6-Gemma2-9B on a 24 GB GPU?

Yes. A GeForce RTX 5090 D V2 with 24 GB runs it at Q8_0, using about 11.6 GB and generating roughly 55.6 tokens per second — a comfortable fit.

Is Ovis1.6-Gemma2-9B open source?

Its weights are published, so Ovis1.6-Gemma2-9B 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 Ovis1.6-Gemma2-9B have?

Ovis1.6-Gemma2-9B has 10.2B parameters. 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 Ovis1.6-Gemma2-9B?

Ovis1.6-Gemma2-9B was published by Alibaba, based in China, categorised as industry.

When was Ovis1.6-Gemma2-9B released?

Ovis1.6-Gemma2-9B was published in September 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 Ovis1.6-Gemma2-9B used for?

Ovis1.6-Gemma2-9B works in Multimodal, Language, Vision, and is recorded as handling language modeling/generation, Visual question answering. Models frequently carry more than one of each, and the tags describe purpose rather than capability limits.

Where can I download Ovis1.6-Gemma2-9B?

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

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