SeamlessM4T TPS calculator

Open weights Facebook,INRIA,University of California (UC) Berkeley 2.3B parameters December 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

818 cards that can run it

818 cards we hold specifications for

Smallest card that fits

Tesla C1080

4 GB · Q8_0 · 16.0 tok/s

Fastest card

B200

1,473 tok/s · 180 GB

Which GPUs can run SeamlessM4T?

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
1,473 tok/s 884–2,357 · low confidence	B200 NVIDIA	180 GB	8,000 GB/s	Jan 2024	3.2 GB	Q8_0	Comfortable
1,473 tok/s 884–2,357 · low confidence	B300 NVIDIA	288 GB	8,000 GB/s	Sep 2025	3.2 GB	Q8_0	Comfortable
1,176 tok/s 706–1,882 · low confidence	Radeon Instinct MI350X AMD	288 GB	8,190 GB/s	Jan 2025	3.2 GB	Q8_0	Comfortable
1,176 tok/s 706–1,882 · low confidence	Radeon Instinct MI355X AMD	288 GB	8,190 GB/s	Jan 2025	3.2 GB	Q8_0	Comfortable
941 tok/s 564–1,505 · low confidence	Radeon Instinct MI300 AMD	128 GB	6,550 GB/s	Jan 2023	3.2 GB	Q8_0	Comfortable
900 tok/s 540–1,441 · low confidence	H200 NVL NVIDIA	141 GB	4,890 GB/s	Nov 2024	3.2 GB	Q8_0	Comfortable
900 tok/s 540–1,441 · low confidence	H200 SXM 141 GB NVIDIA	141 GB	4,890 GB/s	Nov 2024	3.2 GB	Q8_0	Comfortable
862 tok/s 517–1,379 · low confidence	Radeon Instinct MI325X AMD	256 GB	6,000 GB/s	Oct 2024	3.2 GB	Q8_0	Comfortable
765 tok/s 459–1,224 · low confidence	Radeon Instinct MI300A AMD	128 GB	5,325 GB/s	Dec 2023	3.2 GB	Q8_0	Comfortable
765 tok/s 459–1,224 · low confidence	Radeon Instinct MI300X AMD	192 GB	5,325 GB/s	Dec 2023	3.2 GB	Q8_0	Comfortable
765 tok/s 459–1,224 · low confidence	Radeon Instinct MI308X AMD	192 GB	5,325 GB/s	Dec 2023	3.2 GB	Q8_0	Comfortable
726 tok/s 435–1,161 · low confidence	H100 NVL 94 GB NVIDIA	94 GB	3,940 GB/s	Mar 2023	3.2 GB	Q8_0	Comfortable
619 tok/s 371–990 · low confidence	H100 PCIe 96 GB NVIDIA	96 GB	3,360 GB/s	Mar 2023	3.2 GB	Q8_0	Comfortable
619 tok/s 371–990 · low confidence	H100 SXM5 80 GB NVIDIA	80 GB	3,360 GB/s	Oct 2022	3.2 GB	Q8_0	Comfortable
619 tok/s 371–990 · low confidence	H100 SXM5 94 GB NVIDIA	94 GB	3,360 GB/s	Mar 2023	3.2 GB	Q8_0	Comfortable
619 tok/s 371–990 · low confidence	H100 SXM5 96 GB NVIDIA	96 GB	3,360 GB/s	Mar 2023	3.2 GB	Q8_0	Comfortable
619 tok/s 371–990 · low confidence	H800 SXM5 NVIDIA	80 GB	3,360 GB/s	Mar 2023	3.2 GB	Q8_0	Comfortable
471 tok/s 283–754 · low confidence	Radeon Instinct MI250 AMD	128 GB	3,280 GB/s	Nov 2021	3.2 GB	Q8_0	Comfortable
471 tok/s 283–754 · low confidence	Radeon Instinct MI250X AMD	128 GB	3,280 GB/s	Nov 2021	3.2 GB	Q8_0	Comfortable
393 tok/s 236–628 · low confidence	Data Center GPU Max 1550 Intel	128 GB	3,280 GB/s	Jan 2023	3.2 GB	Q8_0	Comfortable
384 tok/s 231–615 · low confidence	Data Center GPU Max Subsystem Intel	128 GB	3,210 GB/s	Jan 2023	3.2 GB	Q8_0	Comfortable
376 tok/s 225–601 · low confidence	A100 SXM4 80 GB NVIDIA	80 GB	2,040 GB/s	Nov 2020	3.2 GB	Q8_0	Comfortable
376 tok/s 225–601 · low confidence	A100X NVIDIA	80 GB	2,040 GB/s	Jun 2021	3.2 GB	Q8_0	Comfortable
376 tok/s 225–601 · low confidence	A800 SXM4 80 GB NVIDIA	80 GB	2,040 GB/s	Aug 2022	3.2 GB	Q8_0	Comfortable
376 tok/s 225–601 · low confidence	H100 CNX NVIDIA	80 GB	2,040 GB/s	Mar 2023	3.2 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: Facebook,INRIA,University of California (UC) Berkeley
Organisation type: Industry,Academia,Academia
Country: United States of America, France
Published: 8 December 2023
Authors: Loïc Barrault, Yu-An Chung, Mariano Coria Meglioli, David Dale, Ning Dong, Mark Duppenthaler, Paul-Ambroise Duquenne, Brian Ellis, Hady Elsahar, Justin Haaheim, John Hoffman, Min-Jae Hwang, Hirofumi Inaguma, Christopher Klaiber, Ilia Kulikov, Pengwei Li, Daniel Licht, Jean Maillard, Ruslan Mavlyutov, Alice Rakotoarison, Kaushik Ram Sadagopan, Abinesh Ramakrishnan, Tuan Tran, Guillaume Wenzek, Yili…

What it does

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

Domain: Speech, Language
Task: Translation, Speech synthesis, Speech recognition (ASR), Speech-to-text, Speech-to-speech
Approach: Self-supervised learning
Base model: W2v-BERT
Numerical format: FP16

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: 2.3B
Training data: tokens

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 V100

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.

Why it is tracked: SOTA improvement
Record confidence: Confident
Citations: 265

Sources

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

Reference: Seamless: Multilingual Expressive and Streaming Speech Translation
Last updated: 25 May 2026

The extremes

The ten fastest GPUs that run SeamlessM4T

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 SeamlessM4T

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

Tesla C1080

Memory needed

3.2 GB

Fastest

1,473 tok/s

SeamlessM4T is small enough at 2.3B parameters that hardware is rarely the obstacle — 818 of the cards we track can run it, including cards several years old.

The entry point is the Tesla C1080: 4 GB of memory, Q8_0 compression, roughly 16.0 tokens per second.

A B200 is the fastest we calculate for it: about 1,473 tokens per second, from 8,000 GB/s of memory bandwidth.

Background

SeamlessM4T was published by Facebook,INRIA,University of California (UC) Berkeley, in United States of America, in December 2023. It comes out of industry,Academia,Academia.

It works in Speech, Language, and is recorded as doing translation, Speech synthesis, Speech recognition (ASR), Speech-to-text, Speech-to-speech.

Its starting point was W2v-BERT — most models at this scale are adapted from an existing base rather than built from nothing.

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.

Reading the throughput figures

The median result is around 41.4 tokens per second; 785 cards produce text faster than most people read it.

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.

Training and provenance

Its inclusion criterion is sOTA improvement.

Step by step

How to choose a GPU for SeamlessM4T

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 SeamlessM4T — around 3.2 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

The conversation occupies memory too, and grows as it goes. Set the slider to the length you expect: at long context SeamlessM4T can slip off a card that handles short questions easily.
03

Choose how far you will compress it

The quantisation column varies by card, because a bigger card holds a more accurate copy of SeamlessM4T — Q8_0 on the smallest card that fits. Set a floor to hold the comparison at one level.
04

Sort by speed

Ranking by tokens per second for SeamlessM4T follows memory bandwidth, not core counts, which is why the B200 tops it at 1,473 tok/s.
05

Look at the headroom, not just the fit

A tight fit runs SeamlessM4T but leaves nothing spare for a longer conversation; comfortable has headroom. If you expect to grow the context, buy for comfortable.
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 SeamlessM4T is settled.

Answers

SeamlessM4T — common questions

Can I run SeamlessM4T on a 8 GB GPU?

Yes. A CMP 170HX 8 GB with 8 GB runs it at Q8_0, using about 3.2 GB and generating roughly 274 tokens per second — a comfortable fit.

Can I run SeamlessM4T on a 12 GB GPU?

Yes. A GeForce RTX 3080 Ti with 12 GB runs it at Q8_0, using about 3.2 GB and generating roughly 168 tokens per second — a comfortable fit.

Can I run SeamlessM4T on a 16 GB GPU?

Yes. A Tesla V100 SXM2 16 GB with 16 GB runs it at Q8_0, using about 3.2 GB and generating roughly 208 tokens per second — a comfortable fit.

Can I run SeamlessM4T on a 24 GB GPU?

Yes. A GeForce RTX 5090 D V2 with 24 GB runs it at Q8_0, using about 3.2 GB and generating roughly 247 tokens per second — a comfortable fit.

Is SeamlessM4T open source?

Its weights are published, so SeamlessM4T 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 SeamlessM4T have?

SeamlessM4T has 2.3B parameters. 2.3B. 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 SeamlessM4T?

SeamlessM4T was published by Facebook,INRIA,University of California (UC) Berkeley, based in United States of America, categorised as industry,Academia,Academia.

When was SeamlessM4T released?

SeamlessM4T was published in December 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.

What is SeamlessM4T used for?

SeamlessM4T works in Speech, Language, and is recorded as handling translation, Speech synthesis, Speech recognition (ASR), Speech-to-text, Speech-to-speech. These are the areas it was designed around; they describe intent rather than a hard boundary.

Where can I download SeamlessM4T?

The weights for SeamlessM4T are published, though we do not hold a repository link for it. This site calculates hardware requirements rather than hosting model files.

Can I run SeamlessM4T if it does not fit in my GPU?

Only by offloading, which is usually a false economy: the part in system memory drags the whole thing down. Our figures for SeamlessM4T assume it is fully resident.

Would two GPUs run SeamlessM4T faster?

Capacity adds across cards; throughput does not. Since 818 of the cards we track already hold SeamlessM4T on their own, a second card is rarely the answer here.

Why does the quantisation differ between cards for SeamlessM4T?

A larger card holds a more accurate copy. Across the cards that run SeamlessM4T, 1 compression levels are used; the floor control above pins it to one.

How accurate are these SeamlessM4T 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 884–2,357 tok/s on the B200 rather than a single number.

What GPU do I need to run SeamlessM4T?

The smallest card in our catalogue that holds SeamlessM4T is the Tesla C1080, with 4 GB of memory. It runs the model at Q8_0 using about 3.2 GB, and produces roughly 16.0 tokens per second. 818 cards in total can run it.

How fast is SeamlessM4T on a GPU?

It depends on the card. The quickest we calculate is a B200 at about 1,473 tokens per second; the slowest that still runs it manages considerably less. Reading speed is around ten tokens per second, and 785 of the cards that can run SeamlessM4T clear that.

How much VRAM does SeamlessM4T need?

About 3.2 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.

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