InternVL1.5 TPS calculator

Open weights Shanghai AI Lab,SenseTime,Tsinghua University,Nanjing University,Fudan University,Chinese University of Hong Kong (CUHK) 25.5B parameters April 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

241 cards that can run it

818 cards we hold specifications for

Smallest card that fits

Xeon Phi 7120P

16 GB · Q3_K_M · 10.3 tok/s

Fastest card

B200

133 tok/s · 180 GB

Which GPUs can run InternVL1.5?

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

241 cards match

Calculating

					Needs	Quantisation	Fit
133 tok/s 80–213 · low confidence	B200 NVIDIA	180 GB	8,000 GB/s	Jan 2024	28.0 GB	Q8_0	Comfortable
133 tok/s 80–213 · low confidence	B300 NVIDIA	288 GB	8,000 GB/s	Sep 2025	28.0 GB	Q8_0	Comfortable
106 tok/s 64–170 · low confidence	Radeon Instinct MI350X AMD	288 GB	8,190 GB/s	Jan 2025	28.0 GB	Q8_0	Comfortable
106 tok/s 64–170 · low confidence	Radeon Instinct MI355X AMD	288 GB	8,190 GB/s	Jan 2025	28.0 GB	Q8_0	Comfortable
84.9 tok/s 51–136 · low confidence	Radeon Instinct MI300 AMD	128 GB	6,550 GB/s	Jan 2023	28.0 GB	Q8_0	Comfortable
81.2 tok/s 49–130 · low confidence	H200 NVL NVIDIA	141 GB	4,890 GB/s	Nov 2024	28.0 GB	Q8_0	Comfortable
81.2 tok/s 49–130 · low confidence	H200 SXM 141 GB NVIDIA	141 GB	4,890 GB/s	Nov 2024	28.0 GB	Q8_0	Comfortable
77.7 tok/s 47–124 · low confidence	Radeon Instinct MI325X AMD	256 GB	6,000 GB/s	Oct 2024	28.0 GB	Q8_0	Comfortable
69.0 tok/s 41–110 · low confidence	Radeon Instinct MI300A AMD	128 GB	5,325 GB/s	Dec 2023	28.0 GB	Q8_0	Comfortable
69.0 tok/s 41–110 · low confidence	Radeon Instinct MI300X AMD	192 GB	5,325 GB/s	Dec 2023	28.0 GB	Q8_0	Comfortable
69.0 tok/s 41–110 · low confidence	Radeon Instinct MI308X AMD	192 GB	5,325 GB/s	Dec 2023	28.0 GB	Q8_0	Comfortable
65.4 tok/s 39–105 · low confidence	H100 NVL 94 GB NVIDIA	94 GB	3,940 GB/s	Mar 2023	28.0 GB	Q8_0	Comfortable
55.8 tok/s 33–89 · low confidence	H100 PCIe 96 GB NVIDIA	96 GB	3,360 GB/s	Mar 2023	28.0 GB	Q8_0	Comfortable
55.8 tok/s 33–89 · low confidence	H100 SXM5 80 GB NVIDIA	80 GB	3,360 GB/s	Oct 2022	28.0 GB	Q8_0	Comfortable
55.8 tok/s 33–89 · low confidence	H100 SXM5 94 GB NVIDIA	94 GB	3,360 GB/s	Mar 2023	28.0 GB	Q8_0	Comfortable
55.8 tok/s 33–89 · low confidence	H100 SXM5 96 GB NVIDIA	96 GB	3,360 GB/s	Mar 2023	28.0 GB	Q8_0	Comfortable
55.8 tok/s 33–89 · low confidence	H800 SXM5 NVIDIA	80 GB	3,360 GB/s	Mar 2023	28.0 GB	Q8_0	Comfortable
50.6 tok/s 30–81 · low confidence	Tesla V100 SXM2 16 GB NVIDIA	16 GB	1,130 GB/s	Nov 2019	13.2 GB	Q3_K_M	Tight
43.0 tok/s 26–69 · low confidence	GeForce RTX 5080 NVIDIA	16 GB	960 GB/s	Jan 2025	13.2 GB	Q3_K_M	Tight
42.5 tok/s 26–68 · low confidence	Radeon Instinct MI250 AMD	128 GB	3,280 GB/s	Nov 2021	28.0 GB	Q8_0	Comfortable
42.5 tok/s 26–68 · low confidence	Radeon Instinct MI250X AMD	128 GB	3,280 GB/s	Nov 2021	28.0 GB	Q8_0	Comfortable
40.2 tok/s 24–64 · low confidence	Tesla V100 DGXS 16 GB NVIDIA	16 GB	897 GB/s	Mar 2018	13.2 GB	Q3_K_M	Tight
40.2 tok/s 24–64 · low confidence	Tesla V100 PCIe 16 GB NVIDIA	16 GB	897 GB/s	Jun 2017	13.2 GB	Q3_K_M	Tight
40.2 tok/s 24–64 · low confidence	Tesla V100 SXM2 16 GB NVIDIA	16 GB	897 GB/s	Jun 2017	13.2 GB	Q3_K_M	Tight
40.2 tok/s 24–64 · low confidence	GeForce RTX 5070 Ti NVIDIA	16 GB	896 GB/s	Feb 2025	13.2 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: Shanghai AI Lab,SenseTime,Tsinghua University,Nanjing University,Fudan University,Chinese University of Hong Kong (CUHK)
Organisation type: Academia,Industry,Academia,Academia,Academia,Academia
Country: China, Hong Kong
Published: 29 April 2024
Authors: Zhe Chen, Weiyun Wang, Hao Tian, Shenglong Ye, Zhangwei Gao, Erfei Cui, Wenwen Tong, Kongzhi Hu, Jiapeng Luo, Zheng Ma, Ji Ma, Jiaqi Wang, Xiaoyi Dong, Hang Yan, Hewei Guo, Conghui He, Botian Shi, Zhenjiang Jin, Chao Xu, Bin Wang, Xingjian Wei, Wei Li, Wenjian Zhang, Bo Zhang, Pinlong Cai, Licheng Wen, Xiangchao Yan, Min Dou, Lewei Lu, Xizhou Zhu, Tong Lu, Dahua Lin, Yu Qiao, Jifeng Dai, Wenhai Wa…

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, Image captioning, Object detection, Character recognition (OCR), Language modeling/generation, Translation
Base model: InternViT-6B,InternLM2-20B

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: 25.5B
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)
Training code: Unreleased
Hugging Face: OpenGVLab

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: How Far Are We to GPT-4V? Closing the Gap to Commercial Multimodal Models with Open-Source Suites
Last updated: 28 November 2025

The extremes

The ten fastest GPUs that run InternVL1.5

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 InternVL1.5

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 7120P

Memory needed

13.2 GB

Fastest

133 tok/s

InternVL1.5 reaches a parameter count of 25.5B. That lands in the range a serious desktop card can handle once the weights are compressed. The number of cards we track that can run it: 241.

At the low end it is handled by Xeon Phi 7120P, with a memory capacity of 16 GB, running it at a compression of Q3_K_M and producing around 10.3 tokens per second.

Top of the range is B200, generating roughly 133 tokens per second on the strength of a memory bandwidth of 8,000 GB/s.

Background

InternVL1.5 was published by Shanghai AI Lab,SenseTime,Tsinghua University,Nanjing University,Fudan University,Chinese University of Hong Kong (CUHK), in the country recorded as China, during April 2024. The category the publisher falls under is academia,Industry,Academia,Academia,Academia,Academia.

It works in the domain of Multimodal, Vision, Language, and is recorded as performing the task of visual question answering, Image captioning, Object detection, Character recognition (OCR), Language modeling/generation, Translation.

Rather than being trained from scratch, it is derived from InternViT-6B,InternLM2-20B. That is why it shares the base model's general shape and size.

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 OpenGVLab.

Reading the throughput figures

Across every card that can run it, the middle of the range sits at 19.9 tokens per second. Exceeding reading speed outright: 195 of them.

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.

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 InternVL1.5

The table above has already assessed every card we hold specifications for against this model. Getting to your answer takes six steps.

01

Check what it needs before anything else

The table lists every card able to hold InternVL1.5, needing around 13.2 GB at a compression of Q3_K_M. That figure, not the headline performance of a card, 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, because at long context a card that handles short questions easily can be dropped by InternVL1.5.
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 InternVL1.5. It will not match a gaming ordering, because generation is bound by memory bandwidth. The card topping the list is B200, at 133 tok/s.
05

Read the fit column last

Tight means it loads and works with no room to raise the context later, in the case of InternVL1.5. 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

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 InternVL1.5.

Answers

InternVL1.5 — common questions

InternVL1.5— 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 Q5_K_M, using about 19.1 GB and generating roughly 39.8 tokens per second. The fit is tight.

InternVL1.5— 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.

InternVL1.5— how many parameters does it have?

It has a parameter count of 25.5B. 25.5B. 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.

InternVL1.5— who created it?

It was published by Shanghai AI Lab,SenseTime,Tsinghua University,Nanjing University,Fudan University,Chinese University of Hong Kong (CUHK), based in China, an organisation categorised as academia,Industry,Academia,Academia,Academia,Academia.

InternVL1.5— when was it released?

It was published in April 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.

InternVL1.5— 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, Image captioning, Object detection, Character recognition (OCR), Language modeling/generation, Translation. These are the areas it was designed around; they describe intent rather than a hard boundary.

InternVL1.5— where can I download it?

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

InternVL1.5— 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 5.3 GB. Every figure here assumes the whole model is resident on the card.

InternVL1.5— would two GPUs run it faster?

A second card roughly doubles the memory available but not the generation rate. The number already able to run it alone: 241. So a second card is rarely the answer here.

InternVL1.5— 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: 5. Bigger cards get the more accurate version, and the quality floor above pins the comparison to one level.

InternVL1.5— how accurate are these speed estimates?

These are estimates with real error bars, and any of them could reasonably land anywhere in its published range depending on which runtime you use. The fastest result here: 80–213 tok/s on B200. The same model and card vary by thirty to fifty per cent depending on the inference software and its version.

InternVL1.5— what GPU do I need to run it?

The smallest card in our catalogue that holds it is Xeon Phi 7120P, with a memory capacity of 16 GB. It runs the model at a compression of Q3_K_M using about 13.2 GB, and produces roughly 10.3 tokens per second. The number of cards able to run it in total: 241.

InternVL1.5— how fast is it on a GPU?

It depends on the card. The quickest we calculate is B200, at about 133 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: 195.

InternVL1.5— how much VRAM does it need?

It needs about 13.2 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.

InternVL1.5— 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 Q3_K_M, using about 13.2 GB and generating roughly 50.6 tokens per second. The fit is tight.

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