Whisper TPS calculator

Open weights OpenAI 1.6B parameters September 2022

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 · 23.8 tok/s

Fastest card

B200

2,186 tok/s · 180 GB

Which GPUs can run Whisper?

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
2,186 tok/s 1,312–3,498 · low confidence	B200 NVIDIA	180 GB	8,000 GB/s	Jan 2024	2.4 GB	Q8_0	Comfortable
2,186 tok/s 1,312–3,498 · low confidence	B300 NVIDIA	288 GB	8,000 GB/s	Sep 2025	2.4 GB	Q8_0	Comfortable
1,746 tok/s 1,047–2,793 · low confidence	Radeon Instinct MI350X AMD	288 GB	8,190 GB/s	Jan 2025	2.4 GB	Q8_0	Comfortable
1,746 tok/s 1,047–2,793 · low confidence	Radeon Instinct MI355X AMD	288 GB	8,190 GB/s	Jan 2025	2.4 GB	Q8_0	Comfortable
1,396 tok/s 838–2,234 · low confidence	Radeon Instinct MI300 AMD	128 GB	6,550 GB/s	Jan 2023	2.4 GB	Q8_0	Comfortable
1,336 tok/s 802–2,138 · low confidence	H200 NVL NVIDIA	141 GB	4,890 GB/s	Nov 2024	2.4 GB	Q8_0	Comfortable
1,336 tok/s 802–2,138 · low confidence	H200 SXM 141 GB NVIDIA	141 GB	4,890 GB/s	Nov 2024	2.4 GB	Q8_0	Comfortable
1,279 tok/s 767–2,046 · low confidence	Radeon Instinct MI325X AMD	256 GB	6,000 GB/s	Oct 2024	2.4 GB	Q8_0	Comfortable
1,135 tok/s 681–1,816 · low confidence	Radeon Instinct MI300A AMD	128 GB	5,325 GB/s	Dec 2023	2.4 GB	Q8_0	Comfortable
1,135 tok/s 681–1,816 · low confidence	Radeon Instinct MI300X AMD	192 GB	5,325 GB/s	Dec 2023	2.4 GB	Q8_0	Comfortable
1,135 tok/s 681–1,816 · low confidence	Radeon Instinct MI308X AMD	192 GB	5,325 GB/s	Dec 2023	2.4 GB	Q8_0	Comfortable
1,077 tok/s 646–1,723 · low confidence	H100 NVL 94 GB NVIDIA	94 GB	3,940 GB/s	Mar 2023	2.4 GB	Q8_0	Comfortable
918 tok/s 551–1,469 · low confidence	H100 PCIe 96 GB NVIDIA	96 GB	3,360 GB/s	Mar 2023	2.4 GB	Q8_0	Comfortable
918 tok/s 551–1,469 · low confidence	H100 SXM5 80 GB NVIDIA	80 GB	3,360 GB/s	Oct 2022	2.4 GB	Q8_0	Comfortable
918 tok/s 551–1,469 · low confidence	H100 SXM5 94 GB NVIDIA	94 GB	3,360 GB/s	Mar 2023	2.4 GB	Q8_0	Comfortable
918 tok/s 551–1,469 · low confidence	H100 SXM5 96 GB NVIDIA	96 GB	3,360 GB/s	Mar 2023	2.4 GB	Q8_0	Comfortable
918 tok/s 551–1,469 · low confidence	H800 SXM5 NVIDIA	80 GB	3,360 GB/s	Mar 2023	2.4 GB	Q8_0	Comfortable
699 tok/s 419–1,119 · low confidence	Radeon Instinct MI250 AMD	128 GB	3,280 GB/s	Nov 2021	2.4 GB	Q8_0	Comfortable
699 tok/s 419–1,119 · low confidence	Radeon Instinct MI250X AMD	128 GB	3,280 GB/s	Nov 2021	2.4 GB	Q8_0	Comfortable
583 tok/s 350–932 · low confidence	Data Center GPU Max 1550 Intel	128 GB	3,280 GB/s	Jan 2023	2.4 GB	Q8_0	Comfortable
570 tok/s 342–912 · low confidence	Data Center GPU Max Subsystem Intel	128 GB	3,210 GB/s	Jan 2023	2.4 GB	Q8_0	Comfortable
557 tok/s 334–892 · low confidence	A100 SXM4 80 GB NVIDIA	80 GB	2,040 GB/s	Nov 2020	2.4 GB	Q8_0	Comfortable
557 tok/s 334–892 · low confidence	A100X NVIDIA	80 GB	2,040 GB/s	Jun 2021	2.4 GB	Q8_0	Comfortable
557 tok/s 334–892 · low confidence	A800 SXM4 80 GB NVIDIA	80 GB	2,040 GB/s	Aug 2022	2.4 GB	Q8_0	Comfortable
557 tok/s 334–892 · low confidence	H100 CNX NVIDIA	80 GB	2,040 GB/s	Mar 2023	2.4 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: OpenAI
Organisation type: Industry
Country: United States of America
Published: 21 September 2022
Authors: Alec Radford, Jong Wook Kim, Tao Xu, Greg Brockman, Christine McLeavey, Ilya Sutskever

What it does

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

Domain: Speech
Task: Speech recognition (ASR)
Approach: Self-supervised learning
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: 1.6B
Training data: 12,403,200,000 tokens
Epochs: 3
Batch size: 256

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: 4.2 × 10²¹ FLOP
How it was established: Hardware

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

How it is classified

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

Foundation model: Yes
Why it is tracked: SOTA improvement
Record confidence: Likely
Citations: 7,066

Sources

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

Reference: Robust Speech Recognition via Large-Scale Weak Supervision
Last updated: 25 May 2026

The extremes

The ten fastest GPUs that run Whisper

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 Whisper

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

What you need to run it

Minimum card

Tesla C1080

Memory needed

2.4 GB

Fastest

2,186 tok/s

Whisper is small enough at 1.6B parameters that hardware is rarely the obstacle — 818 of the cards we track can run it, including cards several years old.

The least hardware that works is a Tesla C1080. Its 4 GB is enough at Q8_0 compression, giving roughly 23.8 tokens per second.

A B200 is the fastest we calculate for it: about 2,186 tokens per second, from 8,000 GB/s of memory bandwidth.

Where it came from

Whisper was published by OpenAI, in United States of America, in September 2022. The organisation is categorised as industry.

It works in Speech, and is recorded as doing speech recognition (ASR).

The weights are published, so it can be downloaded and run on your own hardware indefinitely, offline, with no account attached.

Understanding the speeds

Across every card that can run it, the middle of the range is about 61.4 tokens per second, and 796 of them clear the ten tokens per second that roughly matches reading speed.

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.

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

Training it took roughly 4.2 × 10²¹ FLOP of computation — a measure of what producing the model cost, not of how fast it answers.

It was trained on about 12,403,200,000 tokens of text.

It is tracked in the underlying dataset for one reason in particular: sOTA improvement.

Step by step

How to choose a GPU for Whisper

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

01

Read the memory figure first

Look at what Whisper actually needs — around 2.4 GB at Q8_0. No amount of processing power compensates for a card that cannot hold it.
02

Set the context length you will work at

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

Decide how much compression you will accept

Compression is what makes Whisper 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 Whisper follows memory bandwidth, not core counts, which is why the B200 tops it at 2,186 tok/s.
05

Read the fit column last

A tight fit runs Whisper but leaves nothing spare for a longer conversation; comfortable has headroom. If you expect to grow the context, buy for comfortable.
06

Open the card you have settled on

Every card name links to its own page, which runs the same calculation across the whole model catalogue. Worth a look before buying for Whisper alone — a card is usually bought for more than one model.

Answers

Whisper — common questions

Can I run Whisper on a 16 GB GPU?

Yes. A Tesla V100 SXM2 16 GB with 16 GB runs it at Q8_0, using about 2.4 GB and generating roughly 309 tokens per second — a comfortable fit.

Can I run Whisper on a 24 GB GPU?

Yes. A GeForce RTX 5090 D V2 with 24 GB runs it at Q8_0, using about 2.4 GB and generating roughly 366 tokens per second — a comfortable fit.

Is Whisper open source?

Its weights are published, so Whisper 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 Whisper have?

Whisper has 1.6B parameters. Table 1. 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 Whisper?

Whisper was published by OpenAI, based in United States of America, categorised as industry.

When was Whisper released?

Whisper was published in September 2022. 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 Whisper used for?

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

Where can I download Whisper?

The weights for Whisper 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 Whisper?

Around 4.2 × 10²¹ FLOP. 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 Whisper 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 Whisper is rarely worth using. Every figure here assumes the whole model is on the card.

Would two GPUs run Whisper faster?

Two cards buy memory rather than speed. That matters for Whisper only if one card cannot hold it — 818 can, so a second adds little.

Why does the quantisation differ between cards for Whisper?

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

How accurate are these Whisper speed estimates?

They are calculated from specifications rather than measured, and each carries a range — 1,312–3,498 tok/s on the B200, for instance. The same model and card vary by thirty to fifty per cent depending on the inference software and its version.

What GPU do I need to run Whisper?

The smallest card in our catalogue that holds Whisper is the Tesla C1080, with 4 GB of memory. It runs the model at Q8_0 using about 2.4 GB, and produces roughly 23.8 tokens per second. 818 cards in total can run it.

How fast is Whisper on a GPU?

It depends on the card. The quickest we calculate is a B200 at about 2,186 tokens per second; the slowest that still runs it manages considerably less. Reading speed is around ten tokens per second, and 796 of the cards that can run Whisper clear that.

How much VRAM does Whisper need?

About 2.4 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 Whisper on a 8 GB GPU?

Yes. A CMP 170HX 8 GB with 8 GB runs it at Q8_0, using about 2.4 GB and generating roughly 407 tokens per second — a comfortable fit.

Can I run Whisper on a 12 GB GPU?

Yes. A GeForce RTX 3080 Ti with 12 GB runs it at Q8_0, using about 2.4 GB and generating roughly 249 tokens per second — a comfortable fit.

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