TiTok-L TPS calculator

Open weights ByteDance,Technical University of Munich 307M parameters June 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

818 cards that can run it

818 cards we hold specifications for

Smallest card that fits

Tesla C1080

4 GB · Q8_0 · 120 tok/s

Fastest card

B200

11,037 tok/s · 180 GB

Which GPUs can run TiTok-L?

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

818 cards match

Calculating

					Needs	Quantisation	Fit
11,037 tok/s 6,622–17,659 · low confidence	B200 NVIDIA	180 GB	8,000 GB/s	Jan 2024	1.0 GB	Q8_0	Comfortable
11,037 tok/s 6,622–17,659 · low confidence	B300 NVIDIA	288 GB	8,000 GB/s	Sep 2025	1.0 GB	Q8_0	Comfortable
8,813 tok/s 5,288–14,101 · low confidence	Radeon Instinct MI350X AMD	288 GB	8,190 GB/s	Jan 2025	1.0 GB	Q8_0	Comfortable
8,813 tok/s 5,288–14,101 · low confidence	Radeon Instinct MI355X AMD	288 GB	8,190 GB/s	Jan 2025	1.0 GB	Q8_0	Comfortable
7,048 tok/s 4,229–11,277 · low confidence	Radeon Instinct MI300 AMD	128 GB	6,550 GB/s	Jan 2023	1.0 GB	Q8_0	Comfortable
6,746 tok/s 4,048–10,794 · low confidence	H200 NVL NVIDIA	141 GB	4,890 GB/s	Nov 2024	1.0 GB	Q8_0	Comfortable
6,746 tok/s 4,048–10,794 · low confidence	H200 SXM 141 GB NVIDIA	141 GB	4,890 GB/s	Nov 2024	1.0 GB	Q8_0	Comfortable
6,456 tok/s 3,874–10,330 · low confidence	Radeon Instinct MI325X AMD	256 GB	6,000 GB/s	Oct 2024	1.0 GB	Q8_0	Comfortable
5,730 tok/s 3,438–9,168 · low confidence	Radeon Instinct MI300A AMD	128 GB	5,325 GB/s	Dec 2023	1.0 GB	Q8_0	Comfortable
5,730 tok/s 3,438–9,168 · low confidence	Radeon Instinct MI300X AMD	192 GB	5,325 GB/s	Dec 2023	1.0 GB	Q8_0	Comfortable
5,730 tok/s 3,438–9,168 · low confidence	Radeon Instinct MI308X AMD	192 GB	5,325 GB/s	Dec 2023	1.0 GB	Q8_0	Comfortable
5,436 tok/s 3,261–8,697 · low confidence	H100 NVL 94 GB NVIDIA	94 GB	3,940 GB/s	Mar 2023	1.0 GB	Q8_0	Comfortable
4,635 tok/s 2,781–7,417 · low confidence	H100 PCIe 96 GB NVIDIA	96 GB	3,360 GB/s	Mar 2023	1.0 GB	Q8_0	Comfortable
4,635 tok/s 2,781–7,417 · low confidence	H100 SXM5 80 GB NVIDIA	80 GB	3,360 GB/s	Oct 2022	1.0 GB	Q8_0	Comfortable
4,635 tok/s 2,781–7,417 · low confidence	H100 SXM5 94 GB NVIDIA	94 GB	3,360 GB/s	Mar 2023	1.0 GB	Q8_0	Comfortable
4,635 tok/s 2,781–7,417 · low confidence	H100 SXM5 96 GB NVIDIA	96 GB	3,360 GB/s	Mar 2023	1.0 GB	Q8_0	Comfortable
4,635 tok/s 2,781–7,417 · low confidence	H800 SXM5 NVIDIA	80 GB	3,360 GB/s	Mar 2023	1.0 GB	Q8_0	Comfortable
3,530 tok/s 2,118–5,647 · low confidence	Radeon Instinct MI250 AMD	128 GB	3,280 GB/s	Nov 2021	1.0 GB	Q8_0	Comfortable
3,530 tok/s 2,118–5,647 · low confidence	Radeon Instinct MI250X AMD	128 GB	3,280 GB/s	Nov 2021	1.0 GB	Q8_0	Comfortable
2,941 tok/s 1,765–4,706 · low confidence	Data Center GPU Max 1550 Intel	128 GB	3,280 GB/s	Jan 2023	1.0 GB	Q8_0	Comfortable
2,878 tok/s 1,727–4,606 · low confidence	Data Center GPU Max Subsystem Intel	128 GB	3,210 GB/s	Jan 2023	1.0 GB	Q8_0	Comfortable
2,814 tok/s 1,689–4,503 · low confidence	A100 SXM4 80 GB NVIDIA	80 GB	2,040 GB/s	Nov 2020	1.0 GB	Q8_0	Comfortable
2,814 tok/s 1,689–4,503 · low confidence	A100X NVIDIA	80 GB	2,040 GB/s	Jun 2021	1.0 GB	Q8_0	Comfortable
2,814 tok/s 1,689–4,503 · low confidence	A800 SXM4 80 GB NVIDIA	80 GB	2,040 GB/s	Aug 2022	1.0 GB	Q8_0	Comfortable
2,814 tok/s 1,689–4,503 · low confidence	H100 CNX NVIDIA	80 GB	2,040 GB/s	Mar 2023	1.0 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: ByteDance,Technical University of Munich
Organisation type: Industry,Academia
Country: China, Germany
Published: 11 June 2024
Authors: Qihang Yu, Mark Weber, Xueqing Deng, Xiaohui Shen, Daniel Cremers, Liang-Chieh Chen

What it does

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

Domain: Image generation
Task: Image generation

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: 307M
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.

Training compute: 1.7 × 10²¹ FLOP
How it was established: Hardware

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 SXM4 80 GB
Chip-hours: 5,120

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

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: An Image is Worth 32 Tokens for Reconstruction and Generation
Last updated: 28 November 2025

The extremes

The ten fastest GPUs that run TiTok-L

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 TiTok-L

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

Tesla C1080

Memory needed

1.0 GB

Fastest

11,037 tok/s

TiTok-L is small enough at 307M parameters that hardware is rarely the obstacle — 818 of the cards we track can run it, including cards several years old.

At the low end, a Tesla C1080 handles it — 4 GB, at Q8_0, for about 120 tokens per second.

A B200 is the fastest we calculate for it: about 11,037 tokens per second, from 8,000 GB/s of memory bandwidth.

What this model is

TiTok-L was published by ByteDance,Technical University of Munich, in China, in June 2024. It comes out of industry,Academia.

It works in Image generation, and is recorded as doing image generation.

Because its weights were released, nothing about running it depends on a provider staying available — it is yours once downloaded.

What decides the speed

The median result is around 309.9 tokens per second; 818 cards produce text faster than most people read it.

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.

How it was trained

Producing it required around 1.7 × 10²¹ FLOP of arithmetic, on NVIDIA A100 SXM4 80 GB, which is a statement about the training budget rather than about inference.

Step by step

How to choose a GPU for TiTok-L

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 that can hold TiTok-L — around 1.0 GB at Q8_0. That figure, not the card's headline performance, is what decides whether it runs.
02

Decide how long your conversations run

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

Set a quality floor

Compression is what makes TiTok-L fit smaller cards, at some cost in accuracy — Q8_0 on the smallest card that fits. A minimum quality removes the ones that go too far.
04

Compare tokens per second, not specifications

Ranking by tokens per second for TiTok-L follows memory bandwidth, not core counts, which is why the B200 tops it at 11,037 tok/s.
05

Check the fit verdict before buying

The fit column separates cards that just manage TiTok-L from those with room to spare. Buy for the second if the context might grow.
06

See what else that card runs

Following a card through to its own page shows every other model it can hold, which is the question that follows once TiTok-L is settled.

Answers

TiTok-L — common questions

Why does the quantisation differ between cards for TiTok-L?

Because capacity varies, so does how hard TiTok-L has to be squeezed — 1 distinct levels appear in the table above. Set a minimum quality to compare at one.

How accurate are these TiTok-L 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 6,622–17,659 tok/s on the B200 rather than a single number.

What GPU do I need to run TiTok-L?

The smallest card in our catalogue that holds TiTok-L is the Tesla C1080, with 4 GB of memory. It runs the model at Q8_0 using about 1.0 GB, and produces roughly 120 tokens per second. 818 cards in total can run it.

How fast is TiTok-L on a GPU?

It depends on the card. The quickest we calculate is a B200 at about 11,037 tokens per second; the slowest that still runs it manages considerably less. Reading speed is around ten tokens per second, and 818 of the cards that can run TiTok-L clear that.

How much VRAM does TiTok-L need?

About 1.0 GB at Q8_0 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 TiTok-L on a 8 GB GPU?

Yes. A CMP 170HX 8 GB with 8 GB runs it at Q8_0, using about 1.0 GB and generating roughly 2,056 tokens per second — a comfortable fit.

Can I run TiTok-L on a 12 GB GPU?

Yes. A GeForce RTX 3080 Ti with 12 GB runs it at Q8_0, using about 1.0 GB and generating roughly 1,259 tokens per second — a comfortable fit.

Can I run TiTok-L on a 16 GB GPU?

Yes. A Tesla V100 SXM2 16 GB with 16 GB runs it at Q8_0, using about 1.0 GB and generating roughly 1,559 tokens per second — a comfortable fit.

Can I run TiTok-L on a 24 GB GPU?

Yes. A GeForce RTX 5090 D V2 with 24 GB runs it at Q8_0, using about 1.0 GB and generating roughly 1,849 tokens per second — a comfortable fit.

Is TiTok-L open source?

Its weights are published, so TiTok-L 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 TiTok-L have?

TiTok-L has 307M parameters. "For TiTok variants, we primarily investigate three model sizes—small, base, and large (i.e., TiTok-S, TiTok-B, TiTok-L)—comprising 22M, 86M, and 307M parameters for encoder and decoder, respectively". 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 TiTok-L?

TiTok-L was published by ByteDance,Technical University of Munich, based in China, categorised as industry,Academia.

When was TiTok-L released?

TiTok-L was published in June 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 TiTok-L used for?

TiTok-L works in Image generation, and is recorded as handling image generation. Models frequently carry more than one of each, and the tags describe purpose rather than capability limits.

Where can I download TiTok-L?

The weights for TiTok-L are published, though we do not hold a repository link for it. This site calculates hardware requirements rather than hosting model files.

How much compute was used to train TiTok-L?

Around 1.7 × 10²¹ FLOP, on NVIDIA A100 SXM4 80 GB. 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 TiTok-L 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 TiTok-L is rarely worth using. Every figure here assumes the whole model is on the card.

Would two GPUs run TiTok-L faster?

Capacity adds across cards; throughput does not. Since 818 of the cards we track already hold TiTok-L on their own, a second card is rarely the answer here.

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