OmniFusion-7B (InternViT-6B-448px V1-2) TPS calculator

Open weights AIRI Artificial Intelligence Research Institute,Sber,Skolkovo Institute of Science and Technology 12.5B parameters November 2023

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 · Q3_K_M · 19.0 tok/s

Fastest card

B200

270 tok/s · 180 GB

Which GPUs can run OmniFusion-7B (InternViT-6B-448px V1-2)?

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
270 tok/s 162–432 · low confidence	B200 NVIDIA	180 GB	8,000 GB/s	Jan 2024	14.1 GB	Q8_0	Comfortable
270 tok/s 162–432 · low confidence	B300 NVIDIA	288 GB	8,000 GB/s	Sep 2025	14.1 GB	Q8_0	Comfortable
216 tok/s 129–345 · low confidence	Radeon Instinct MI350X AMD	288 GB	8,190 GB/s	Jan 2025	14.1 GB	Q8_0	Comfortable
216 tok/s 129–345 · low confidence	Radeon Instinct MI355X AMD	288 GB	8,190 GB/s	Jan 2025	14.1 GB	Q8_0	Comfortable
173 tok/s 104–276 · low confidence	Radeon Instinct MI300 AMD	128 GB	6,550 GB/s	Jan 2023	14.1 GB	Q8_0	Comfortable
165 tok/s 99–264 · low confidence	H200 NVL NVIDIA	141 GB	4,890 GB/s	Nov 2024	14.1 GB	Q8_0	Comfortable
165 tok/s 99–264 · low confidence	H200 SXM 141 GB NVIDIA	141 GB	4,890 GB/s	Nov 2024	14.1 GB	Q8_0	Comfortable
158 tok/s 95–253 · low confidence	Radeon Instinct MI325X AMD	256 GB	6,000 GB/s	Oct 2024	14.1 GB	Q8_0	Comfortable
140 tok/s 84–224 · low confidence	Radeon Instinct MI300A AMD	128 GB	5,325 GB/s	Dec 2023	14.1 GB	Q8_0	Comfortable
140 tok/s 84–224 · low confidence	Radeon Instinct MI300X AMD	192 GB	5,325 GB/s	Dec 2023	14.1 GB	Q8_0	Comfortable
140 tok/s 84–224 · low confidence	Radeon Instinct MI308X AMD	192 GB	5,325 GB/s	Dec 2023	14.1 GB	Q8_0	Comfortable
136 tok/s 81–217 · low confidence	CMP 170HX 8 GB NVIDIA	8 GB	1,490 GB/s	Sep 2021	6.8 GB	Q3_K_M	Tight
133 tok/s 80–213 · low confidence	H100 NVL 94 GB NVIDIA	94 GB	3,940 GB/s	Mar 2023	14.1 GB	Q8_0	Comfortable
122 tok/s 73–195 · low confidence	CMP 170HX 10 GB NVIDIA	10 GB	1,560 GB/s	Sep 2021	8.3 GB	Q4_K_M	Tight
113 tok/s 68–182 · low confidence	H100 PCIe 96 GB NVIDIA	96 GB	3,360 GB/s	Mar 2023	14.1 GB	Q8_0	Comfortable
113 tok/s 68–182 · low confidence	H100 SXM5 80 GB NVIDIA	80 GB	3,360 GB/s	Oct 2022	14.1 GB	Q8_0	Comfortable
113 tok/s 68–182 · low confidence	H100 SXM5 94 GB NVIDIA	94 GB	3,360 GB/s	Mar 2023	14.1 GB	Q8_0	Comfortable
113 tok/s 68–182 · low confidence	H100 SXM5 96 GB NVIDIA	96 GB	3,360 GB/s	Mar 2023	14.1 GB	Q8_0	Comfortable
113 tok/s 68–182 · low confidence	H800 SXM5 NVIDIA	80 GB	3,360 GB/s	Mar 2023	14.1 GB	Q8_0	Comfortable
86.4 tok/s 52–138 · low confidence	Radeon Instinct MI250 AMD	128 GB	3,280 GB/s	Nov 2021	14.1 GB	Q8_0	Comfortable
86.4 tok/s 52–138 · low confidence	Radeon Instinct MI250X AMD	128 GB	3,280 GB/s	Nov 2021	14.1 GB	Q8_0	Comfortable
72.0 tok/s 43–115 · low confidence	Data Center GPU Max 1550 Intel	128 GB	3,280 GB/s	Jan 2023	14.1 GB	Q8_0	Comfortable
70.5 tok/s 42–113 · low confidence	Data Center GPU Max Subsystem Intel	128 GB	3,210 GB/s	Jan 2023	14.1 GB	Q8_0	Comfortable
70.0 tok/s 42–112 · low confidence	RTX A5000-8Q NVIDIA	8 GB	768 GB/s	Apr 2021	6.8 GB	Q3_K_M	Tight
68.9 tok/s 41–110 · low confidence	A100 SXM4 80 GB NVIDIA	80 GB	2,040 GB/s	Nov 2020	14.1 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: AIRI Artificial Intelligence Research Institute,Sber,Skolkovo Institute of Science and Technology
Organisation type: Research collective,Industry,Government,Academia
Country: Russia
Published: 22 November 2023
Authors: Elizaveta Goncharova, Anton Razzhigaev, Matvey Mikhalchuk, Maxim Kurkin, Irina Abdullaeva, Matvey Skripkin, Ivan Oseledets, Denis Dimitrov, Andrey Kuznetsov

What it does

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

Domain: Multimodal, Vision, Language
Task: Visual question answering, Language modeling/generation, Question answering
Base model: InternViT-6B,GigaChat-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: 12.5B
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.

Fine-tuning compute: 4 × 10¹⁶ FLOP

The training run

What it physically took to train: which chips, how many, for how long, and what that drew from the wall.

Training hardware: NVIDIA A100
Chips used: 8
Power draw: 6.3 kW

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: Open source
Hugging Face: AIRI-Institute

How it is classified

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

Record confidence: Confident
Citations: 7

Sources

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

Reference: OmniFusion Technical Report
Last updated: 25 May 2026

The extremes

The ten fastest GPUs that run OmniFusion-7B (InternViT-6B-448px V1-2)

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 OmniFusion-7B (InternViT-6B-448px V1-2)

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

Hardware requirements in practice

Minimum card

Xeon Phi 5110P

Memory needed

6.8 GB

Fastest

270 tok/s

OmniFusion-7B (InternViT-6B-448px V1-2) reaches a parameter count of 12.5B. That is small enough that hardware is rarely the obstacle, including on cards several years old. The number of cards we track that can run it: 509.

The entry point is Xeon Phi 5110P, with a memory capacity of 8 GB, running it at a compression of Q3_K_M and producing around 19.0 tokens per second.

The fastest we calculate for it is B200, generating roughly 270 tokens per second on the strength of a memory bandwidth of 8,000 GB/s.

What this model is

OmniFusion-7B (InternViT-6B-448px V1-2) was published by AIRI Artificial Intelligence Research Institute,Sber,Skolkovo Institute of Science and Technology, in the country recorded as Russia, during November 2023. The publishing organisation is categorised as research collective,Industry,Government,Academia.

It works in the domain of Multimodal, Vision, Language, and is recorded as performing the task of visual question answering, Language modeling/generation, Question answering.

Its starting point was an existing base model, InternViT-6B,GigaChat-7B. Most models at this scale are adapted from an existing base rather than built from nothing.

Because its weights were released, nothing about running it depends on a provider staying available — it is yours once downloaded. On Hugging Face it is published under the organisation AIRI-Institute.

What decides the speed

Across every card that can run it, the middle of the range sits at 20.3 tokens per second. Producing text faster than most people read it: 451 of them.

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.

Its internal architecture is not on file, so memory is approximated from the parameter count and marked accordingly. Expect the real figure to differ, more so at long context.

Step by step

How to choose a GPU for OmniFusion-7B (InternViT-6B-448px V1-2)

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

The table lists every card able to hold OmniFusion-7B (InternViT-6B-448px V1-2), needing around 6.8 GB at a compression of 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 OmniFusion-7B (InternViT-6B-448px V1-2).
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

Rank by throughput rather than spec sheet

Sort by speed to see how cards rank for OmniFusion-7B (InternViT-6B-448px V1-2). It will not match a gaming ordering, because generation is bound by memory bandwidth. The card topping the list is B200, at 270 tok/s.
05

Read the fit column last

The fit column separates cards that just manage it from those with room to spare, in the case of OmniFusion-7B (InternViT-6B-448px V1-2). 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

See what else that card runs

Every card name links to its own page, which runs the same calculation across the whole model catalogue. A card is usually bought for more than one model, so it is worth a look before buying for OmniFusion-7B (InternViT-6B-448px V1-2).

Answers

OmniFusion-7B (InternViT-6B-448px V1-2) — common questions

OmniFusion-7B (InternViT-6B-448px V1-2)— 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: 162–432 tok/s on B200. The same model and card vary by thirty to fifty per cent depending on the inference software and its version.

OmniFusion-7B (InternViT-6B-448px V1-2)— what GPU do I need to run it?

The smallest card in our catalogue that holds it is Xeon Phi 5110P, with a memory capacity of 8 GB. It runs the model at a compression of Q3_K_M using about 6.8 GB, and produces roughly 19.0 tokens per second. The number of cards able to run it in total: 509.

OmniFusion-7B (InternViT-6B-448px V1-2)— how fast is it on a GPU?

It depends on the card. The quickest we calculate is B200, at about 270 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: 451.

OmniFusion-7B (InternViT-6B-448px V1-2)— how much VRAM does it need?

It needs about 6.8 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.

OmniFusion-7B (InternViT-6B-448px V1-2)— can I run it on a GPU holding 8 GB?

Yes. The card CMP 170HX 8 GB, holding 8 GB, runs it at a compression of Q3_K_M, using about 6.8 GB and generating roughly 136 tokens per second. The fit is tight.

OmniFusion-7B (InternViT-6B-448px V1-2)— can I run it on a GPU holding 12 GB?

Yes. The card GeForce RTX 3080 Ti, holding 12 GB, runs it at a compression of Q5_K_M, using about 9.7 GB and generating roughly 55.0 tokens per second. The fit is tight.

OmniFusion-7B (InternViT-6B-448px V1-2)— can I run it on a GPU holding 16 GB?

Yes. The card Tesla V100 SXM2 16 GB, holding 16 GB, runs it at a compression of Q8_0, using about 14.1 GB and generating roughly 38.2 tokens per second. The fit is tight.

OmniFusion-7B (InternViT-6B-448px V1-2)— can I run it on a GPU holding 24 GB?

Yes. The card GeForce RTX 5090 D V2, holding 24 GB, runs it at a compression of Q8_0, using about 14.1 GB and generating roughly 45.3 tokens per second. The fit is comfortable.

OmniFusion-7B (InternViT-6B-448px V1-2)— 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.

OmniFusion-7B (InternViT-6B-448px V1-2)— how many parameters does it have?

It has a parameter count of 12.5B. The model uses InternViT-6B-448px V1-2 as the visual encoder, which undergoes pre-training, and GigaChat-7B as the LLM, which undergoes fine-tuning [1]. InternViT-6B-448px V1-2 has 5.54B parameters [2], and GigaChat-7B, being closed-source is assumed to have 7B parameters. 1. https://arxiv.org/pdf/2404.06212 2. https://huggingface.co/OpenGVLab/InternViT-6B-448px-V1-2. 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.

OmniFusion-7B (InternViT-6B-448px V1-2)— who created it?

It was published by AIRI Artificial Intelligence Research Institute,Sber,Skolkovo Institute of Science and Technology, based in Russia, an organisation categorised as research collective,Industry,Government,Academia.

OmniFusion-7B (InternViT-6B-448px V1-2)— when was it released?

It was published in November 2023. Capability per parameter has improved considerably since, so a newer model of the same size is often the better use of the same hardware.

OmniFusion-7B (InternViT-6B-448px V1-2)— what is it used for?

It works in the domain of Multimodal, Vision, Language, and is recorded as handling the task of visual question answering, Language modeling/generation, Question answering. These are the areas it was designed around; they describe intent rather than a hard boundary.

OmniFusion-7B (InternViT-6B-448px V1-2)— where can I download it?

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

OmniFusion-7B (InternViT-6B-448px V1-2)— can I run it if it does not fit in my GPU?

Partly. Layers that do not fit sit in system memory and run at a fraction of the speed, so a mostly-offloaded model is rarely worth using. The nearest miss we calculate falls short by 2.9 GB. Every figure here assumes the whole model is resident on the card.

OmniFusion-7B (InternViT-6B-448px V1-2)— would two GPUs run it faster?

Two cards buy memory rather than speed, which matters only if one card cannot hold it. The number that can: 509. So a second card is rarely the answer here.

OmniFusion-7B (InternViT-6B-448px V1-2)— why does the quantisation differ between cards?

A larger card holds a more accurate copy. The number of compression levels used across the cards that run it: 4. Bigger cards get the more accurate version, and the quality floor above pins the comparison to one level.

Source

Original publication

Record last updated 25 May 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