F5-TTS TPS calculator

Open weights Shanghai Jiao Tong University,University of Cambridge,Geely Automobile Research Institute (Ningbo) Company 335.8M parameters December 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 · 110 tok/s

Fastest card

B200

10,090 tok/s · 180 GB

Which GPUs can run F5-TTS?

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
10,090 tok/s 6,054–16,144 · low confidence	B200 NVIDIA	180 GB	8,000 GB/s	Jan 2024	1.1 GB	Q8_0	Comfortable
10,090 tok/s 6,054–16,144 · low confidence	B300 NVIDIA	288 GB	8,000 GB/s	Sep 2025	1.1 GB	Q8_0	Comfortable
8,057 tok/s 4,834–12,891 · low confidence	Radeon Instinct MI350X AMD	288 GB	8,190 GB/s	Jan 2025	1.1 GB	Q8_0	Comfortable
8,057 tok/s 4,834–12,891 · low confidence	Radeon Instinct MI355X AMD	288 GB	8,190 GB/s	Jan 2025	1.1 GB	Q8_0	Comfortable
6,444 tok/s 3,866–10,310 · low confidence	Radeon Instinct MI300 AMD	128 GB	6,550 GB/s	Jan 2023	1.1 GB	Q8_0	Comfortable
6,168 tok/s 3,701–9,868 · low confidence	H200 NVL NVIDIA	141 GB	4,890 GB/s	Nov 2024	1.1 GB	Q8_0	Comfortable
6,168 tok/s 3,701–9,868 · low confidence	H200 SXM 141 GB NVIDIA	141 GB	4,890 GB/s	Nov 2024	1.1 GB	Q8_0	Comfortable
5,903 tok/s 3,542–9,444 · low confidence	Radeon Instinct MI325X AMD	256 GB	6,000 GB/s	Oct 2024	1.1 GB	Q8_0	Comfortable
5,239 tok/s 3,143–8,382 · low confidence	Radeon Instinct MI300A AMD	128 GB	5,325 GB/s	Dec 2023	1.1 GB	Q8_0	Comfortable
5,239 tok/s 3,143–8,382 · low confidence	Radeon Instinct MI300X AMD	192 GB	5,325 GB/s	Dec 2023	1.1 GB	Q8_0	Comfortable
5,239 tok/s 3,143–8,382 · low confidence	Radeon Instinct MI308X AMD	192 GB	5,325 GB/s	Dec 2023	1.1 GB	Q8_0	Comfortable
4,969 tok/s 2,982–7,951 · low confidence	H100 NVL 94 GB NVIDIA	94 GB	3,940 GB/s	Mar 2023	1.1 GB	Q8_0	Comfortable
4,238 tok/s 2,543–6,781 · low confidence	H100 PCIe 96 GB NVIDIA	96 GB	3,360 GB/s	Mar 2023	1.1 GB	Q8_0	Comfortable
4,238 tok/s 2,543–6,781 · low confidence	H100 SXM5 80 GB NVIDIA	80 GB	3,360 GB/s	Oct 2022	1.1 GB	Q8_0	Comfortable
4,238 tok/s 2,543–6,781 · low confidence	H100 SXM5 94 GB NVIDIA	94 GB	3,360 GB/s	Mar 2023	1.1 GB	Q8_0	Comfortable
4,238 tok/s 2,543–6,781 · low confidence	H100 SXM5 96 GB NVIDIA	96 GB	3,360 GB/s	Mar 2023	1.1 GB	Q8_0	Comfortable
4,238 tok/s 2,543–6,781 · low confidence	H800 SXM5 NVIDIA	80 GB	3,360 GB/s	Mar 2023	1.1 GB	Q8_0	Comfortable
3,227 tok/s 1,936–5,163 · low confidence	Radeon Instinct MI250 AMD	128 GB	3,280 GB/s	Nov 2021	1.1 GB	Q8_0	Comfortable
3,227 tok/s 1,936–5,163 · low confidence	Radeon Instinct MI250X AMD	128 GB	3,280 GB/s	Nov 2021	1.1 GB	Q8_0	Comfortable
2,689 tok/s 1,613–4,302 · low confidence	Data Center GPU Max 1550 Intel	128 GB	3,280 GB/s	Jan 2023	1.1 GB	Q8_0	Comfortable
2,632 tok/s 1,579–4,211 · low confidence	Data Center GPU Max Subsystem Intel	128 GB	3,210 GB/s	Jan 2023	1.1 GB	Q8_0	Comfortable
2,573 tok/s 1,544–4,117 · low confidence	A100 SXM4 80 GB NVIDIA	80 GB	2,040 GB/s	Nov 2020	1.1 GB	Q8_0	Comfortable
2,573 tok/s 1,544–4,117 · low confidence	A100X NVIDIA	80 GB	2,040 GB/s	Jun 2021	1.1 GB	Q8_0	Comfortable
2,573 tok/s 1,544–4,117 · low confidence	A800 SXM4 80 GB NVIDIA	80 GB	2,040 GB/s	Aug 2022	1.1 GB	Q8_0	Comfortable
2,573 tok/s 1,544–4,117 · low confidence	H100 CNX NVIDIA	80 GB	2,040 GB/s	Mar 2023	1.1 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: Shanghai Jiao Tong University,University of Cambridge,Geely Automobile Research Institute (Ningbo) Company
Organisation type: Academia,Academia,Industry
Country: China, United Kingdom of Great Britain and Northern Ireland
Published: 15 December 2024
Authors: Yushen Chen, Zhikang Niu, Ziyang Ma, Keqi Deng, Chunhui Wang, Jian Zhao, Kai Yu, Xie Chen

What it does

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

Domain: Speech
Task: Speech synthesis, Translation

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: 335.8M
Training data: 27,325,800,000 tokens
Batch size: 307,200

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: 4.5 × 10²⁰ 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
Chips used: 8
Wall-clock time: 168 hours (7 days)
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 (non-commercial)
Training code: Open source
Hugging Face: SWivid

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: F5-TTS: A Fairytaler that Fakes Fluent and Faithful Speech with Flow Matching
Last updated: 28 November 2025

The extremes

The ten fastest GPUs that run F5-TTS

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 F5-TTS

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

1.1 GB

Fastest

10,090 tok/s

F5-TTS reaches a parameter count of 335.8M. 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: 818.

The entry point is Tesla C1080, with a memory capacity of 4 GB, running it at a compression of Q8_0 and producing around 110 tokens per second.

At the other end sits B200, generating roughly 10,090 tokens per second on the strength of a memory bandwidth of 8,000 GB/s.

Background

F5-TTS was published by Shanghai Jiao Tong University,University of Cambridge,Geely Automobile Research Institute (Ningbo) Company, in the country recorded as China, during December 2024. The publishing organisation is categorised as academia,Academia,Industry.

It works in the domain of Speech, and is recorded as performing the task of speech synthesis, Translation.

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 SWivid.

Reading the throughput figures

The median result is around 283.3 tokens per second. Exceeding reading speed outright: 818 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.

How it was trained

The training run consumed about 4.5 × 10²⁰ FLOP, on hardware recorded as NVIDIA A100 SXM4 80 GB. That figure measures what producing the model cost, and has no bearing on how fast it answers.

Training consumed a corpus of around 27,325,800,000 tokens of text.

Step by step

How to choose a GPU for F5-TTS

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

Every card here has been checked against F5-TTS, needing around 1.1 GB at a compression of Q8_0. That figure, not the headline performance of a card, is what decides whether it runs.
02

Set the context length you will work at

Longer conversations cost memory on top of what the weights need. Move the slider to your real working length before trusting any row for F5-TTS.
03

Decide how much compression you will accept

Each card runs the least-compressed copy it can hold, reaching a compression of Q8_0 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

Compare tokens per second, not specifications

Ranking by tokens per second follows memory bandwidth rather than core counts, for F5-TTS. It will not match a gaming ordering, because generation is bound by memory bandwidth. The card topping the list is B200, at 10,090 tok/s.
05

Check the fit verdict before buying

A tight fit runs, but leaves nothing spare for a longer conversation, in the case of F5-TTS. 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 F5-TTS.

Answers

F5-TTS — common questions

F5-TTS— how fast is it on a GPU?

It depends on the card. The quickest we calculate is B200, at about 10,090 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: 818.

F5-TTS— how much VRAM does it need?

It needs about 1.1 GB at a compression of Q8_0, 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.

F5-TTS— 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 Q8_0, using about 1.1 GB and generating roughly 1,879 tokens per second. The fit is comfortable.

F5-TTS— 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 Q8_0, using about 1.1 GB and generating roughly 1,151 tokens per second. The fit is comfortable.

F5-TTS— 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 1.1 GB and generating roughly 1,425 tokens per second. The fit is comfortable.

F5-TTS— 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 1.1 GB and generating roughly 1,690 tokens per second. The fit is comfortable.

F5-TTS— 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.

F5-TTS— how many parameters does it have?

It has a parameter count of 335.8M. The F5-TTS base model has 22 layers, 16 attention heads, 1024/2048 embedding/feed-forward network (FFN) dimension for DiT; and 4 layers, 512/1024 embedding/FFN dimension for ConvNeXt V2; in total 335.8M 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.

F5-TTS— who created it?

It was published by Shanghai Jiao Tong University,University of Cambridge,Geely Automobile Research Institute (Ningbo) Company, based in China, an organisation categorised as academia,Academia,Industry.

F5-TTS— when was it released?

It was published in December 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.

F5-TTS— what is it used for?

It works in the domain of Speech, and is recorded as handling the task of speech synthesis, Translation. A model can carry several of each, so these are the areas it was built for rather than a limit on what it will attempt.

F5-TTS— where can I download it?

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

F5-TTS— how much compute was used to train it?

Training consumed around 4.5 × 10²⁰ FLOP, on hardware recorded as 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.

F5-TTS— 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. Every figure here assumes the whole model is resident on the card.

F5-TTS— would two GPUs run it faster?

Capacity adds across cards; throughput does not. The number of cards already holding it on their own: 818. So a second card is rarely the answer here.

F5-TTS— why does the quantisation differ between cards?

Each card is shown running the least-compressed copy it can hold. The number of distinct compression levels across the cards that fit it: 1. Bigger cards get the more accurate version, and the quality floor above pins the comparison to one level.

F5-TTS— how accurate are these speed estimates?

They are calculated from specifications rather than measured, and each carries a range. One example: 6,054–16,144 tok/s on B200. The same model and card vary by thirty to fifty per cent depending on the inference software and its version.

F5-TTS— what GPU do I need to run it?

The smallest card in our catalogue that holds it is Tesla C1080, with a memory capacity of 4 GB. It runs the model at a compression of Q8_0 using about 1.1 GB, and produces roughly 110 tokens per second. The number of cards able to run it in total: 818.

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