Bird Vocalization Classifier (Perch) TPS calculator

Open weights Google Research,Cornell University,Naturalis Biodiversity Center,Chemnitz University of Technology 7.8M 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 · 4,726 tok/s

Fastest card

B200

434,389 tok/s · 180 GB

Which GPUs can run Bird Vocalization Classifier (Perch)?

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
434,389 tok/s 260,633–695,023 · low confidence	B200 NVIDIA	180 GB	8,000 GB/s	Jan 2024	0.7 GB	Q8_0	Comfortable
434,389 tok/s 260,633–695,023 · low confidence	B300 NVIDIA	288 GB	8,000 GB/s	Sep 2025	0.7 GB	Q8_0	Comfortable
346,871 tok/s 208,122–554,993 · low confidence	Radeon Instinct MI350X AMD	288 GB	8,190 GB/s	Jan 2025	0.7 GB	Q8_0	Comfortable
346,871 tok/s 208,122–554,993 · low confidence	Radeon Instinct MI355X AMD	288 GB	8,190 GB/s	Jan 2025	0.7 GB	Q8_0	Comfortable
277,412 tok/s 166,447–443,859 · low confidence	Radeon Instinct MI300 AMD	128 GB	6,550 GB/s	Jan 2023	0.7 GB	Q8_0	Comfortable
265,520 tok/s 159,312–424,833 · low confidence	H200 NVL NVIDIA	141 GB	4,890 GB/s	Nov 2024	0.7 GB	Q8_0	Comfortable
265,520 tok/s 159,312–424,833 · low confidence	H200 SXM 141 GB NVIDIA	141 GB	4,890 GB/s	Nov 2024	0.7 GB	Q8_0	Comfortable
254,118 tok/s 152,471–406,588 · low confidence	Radeon Instinct MI325X AMD	256 GB	6,000 GB/s	Oct 2024	0.7 GB	Q8_0	Comfortable
225,529 tok/s 135,318–360,847 · low confidence	Radeon Instinct MI300A AMD	128 GB	5,325 GB/s	Dec 2023	0.7 GB	Q8_0	Comfortable
225,529 tok/s 135,318–360,847 · low confidence	Radeon Instinct MI300X AMD	192 GB	5,325 GB/s	Dec 2023	0.7 GB	Q8_0	Comfortable
225,529 tok/s 135,318–360,847 · low confidence	Radeon Instinct MI308X AMD	192 GB	5,325 GB/s	Dec 2023	0.7 GB	Q8_0	Comfortable
213,937 tok/s 128,362–342,299 · low confidence	H100 NVL 94 GB NVIDIA	94 GB	3,940 GB/s	Mar 2023	0.7 GB	Q8_0	Comfortable
182,443 tok/s 109,466–291,910 · low confidence	H100 PCIe 96 GB NVIDIA	96 GB	3,360 GB/s	Mar 2023	0.7 GB	Q8_0	Comfortable
182,443 tok/s 109,466–291,910 · low confidence	H100 SXM5 80 GB NVIDIA	80 GB	3,360 GB/s	Oct 2022	0.7 GB	Q8_0	Comfortable
182,443 tok/s 109,466–291,910 · low confidence	H100 SXM5 94 GB NVIDIA	94 GB	3,360 GB/s	Mar 2023	0.7 GB	Q8_0	Comfortable
182,443 tok/s 109,466–291,910 · low confidence	H100 SXM5 96 GB NVIDIA	96 GB	3,360 GB/s	Mar 2023	0.7 GB	Q8_0	Comfortable
182,443 tok/s 109,466–291,910 · low confidence	H800 SXM5 NVIDIA	80 GB	3,360 GB/s	Mar 2023	0.7 GB	Q8_0	Comfortable
138,918 tok/s 83,351–222,268 · low confidence	Radeon Instinct MI250 AMD	128 GB	3,280 GB/s	Nov 2021	0.7 GB	Q8_0	Comfortable
138,918 tok/s 83,351–222,268 · low confidence	Radeon Instinct MI250X AMD	128 GB	3,280 GB/s	Nov 2021	0.7 GB	Q8_0	Comfortable
115,765 tok/s 69,459–185,224 · low confidence	Data Center GPU Max 1550 Intel	128 GB	3,280 GB/s	Jan 2023	0.7 GB	Q8_0	Comfortable
113,294 tok/s 67,976–181,271 · low confidence	Data Center GPU Max Subsystem Intel	128 GB	3,210 GB/s	Jan 2023	0.7 GB	Q8_0	Comfortable
110,769 tok/s 66,462–177,231 · low confidence	A100 SXM4 80 GB NVIDIA	80 GB	2,040 GB/s	Nov 2020	0.7 GB	Q8_0	Comfortable
110,769 tok/s 66,462–177,231 · low confidence	A100X NVIDIA	80 GB	2,040 GB/s	Jun 2021	0.7 GB	Q8_0	Comfortable
110,769 tok/s 66,462–177,231 · low confidence	A800 SXM4 80 GB NVIDIA	80 GB	2,040 GB/s	Aug 2022	0.7 GB	Q8_0	Comfortable
110,769 tok/s 66,462–177,231 · low confidence	H100 CNX NVIDIA	80 GB	2,040 GB/s	Mar 2023	0.7 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: Google Research,Cornell University,Naturalis Biodiversity Center,Chemnitz University of Technology
Organisation type: Industry,Academia,Government,Academia
Country: United States of America, Germany
Published: 18 December 2023
Authors: Burooj Ghani, Tom Denton, Stefan Kahl, Holger Klinck

What it does

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

Domain: Audio
Task: Audio classification
Base model: EfficientNet-B1

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: 7.8M
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: 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: Likely

Sources

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

Reference: Global birdsong embeddings enable superior transfer learning for bioacoustic classification
Last updated: 11 February 2026

The extremes

The ten fastest GPUs that run Bird Vocalization Classifier (Perch)

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 Bird Vocalization Classifier (Perch)

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

0.7 GB

Fastest

434,389 tok/s

Bird Vocalization Classifier (Perch) is small enough at 7.8M 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 4,726 tokens per second.

The quickest result comes from a B200 at around 434,389 tokens per second — its 8,000 GB/s of bandwidth is what buys that.

About this model

Bird Vocalization Classifier (Perch) was published by Google Research,Cornell University,Naturalis Biodiversity Center,Chemnitz University of Technology, in United States of America, in December 2023. industry,Academia,Government,Academia is the category the publisher falls under.

It works in Audio, and is recorded as doing audio classification.

It is derived from EfficientNet-B1 rather than trained from scratch, which is the usual way a specialised model is produced.

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

How fast it runs, and why

Half the cards that hold it manage more than 12,197.7 tokens per second, and 818 exceed reading speed outright.

Every weight participates in every token here, so bandwidth is the whole story: the ranking below is effectively a ranking of memory throughput.

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 Bird Vocalization Classifier (Perch)

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 Bird Vocalization Classifier (Perch) — around 0.7 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 Bird Vocalization Classifier (Perch) can slip off a card that handles short questions easily.
03

Decide how much compression you will accept

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

Compare tokens per second, not specifications

The speed ordering for Bird Vocalization Classifier (Perch) is effectively an ordering by memory bandwidth, which is why the B200 tops it at 434,389 tok/s.
05

Read the fit column last

The fit column separates cards that just manage Bird Vocalization Classifier (Perch) from those with room to spare. Buy for the second if the context might grow.
06

Check the card from the other side

Each card page repeats this sweep for every model we hold. It answers what else the hardware is good for, beyond Bird Vocalization Classifier (Perch).

Answers

Bird Vocalization Classifier (Perch) — common questions

Can I run Bird Vocalization Classifier (Perch) on a 12 GB GPU?

Yes. A GeForce RTX 3080 Ti with 12 GB runs it at Q8_0, using about 0.7 GB and generating roughly 49,542 tokens per second — a comfortable fit.

Can I run Bird Vocalization Classifier (Perch) on a 16 GB GPU?

Yes. A Tesla V100 SXM2 16 GB with 16 GB runs it at Q8_0, using about 0.7 GB and generating roughly 61,357 tokens per second — a comfortable fit.

Can I run Bird Vocalization Classifier (Perch) on a 24 GB GPU?

Yes. A GeForce RTX 5090 D V2 with 24 GB runs it at Q8_0, using about 0.7 GB and generating roughly 72,760 tokens per second — a comfortable fit.

Is Bird Vocalization Classifier (Perch) open source?

Its weights are published, so Bird Vocalization Classifier (Perch) 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 Bird Vocalization Classifier (Perch) have?

Bird Vocalization Classifier (Perch) has 7.8M parameters. EfficientNet-B1 architecture (7.8M parameters) - I assume this model has the same amount embedding size: 1280. 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 Bird Vocalization Classifier (Perch)?

Bird Vocalization Classifier (Perch) was published by Google Research,Cornell University,Naturalis Biodiversity Center,Chemnitz University of Technology, based in United States of America, categorised as industry,Academia,Government,Academia.

When was Bird Vocalization Classifier (Perch) released?

Bird Vocalization Classifier (Perch) 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 Bird Vocalization Classifier (Perch) used for?

Bird Vocalization Classifier (Perch) works in Audio, and is recorded as handling audio classification. Models frequently carry more than one of each, and the tags describe purpose rather than capability limits.

Where can I download Bird Vocalization Classifier (Perch)?

The weights for Bird Vocalization Classifier (Perch) 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 Bird Vocalization Classifier (Perch) 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 Bird Vocalization Classifier (Perch) is rarely worth using. Every figure here assumes the whole model is on the card.

Would two GPUs run Bird Vocalization Classifier (Perch) faster?

A second card roughly doubles the memory available but not the generation rate. With 818 cards already able to run Bird Vocalization Classifier (Perch) alone, the case for pairing is weak.

Why does the quantisation differ between cards for Bird Vocalization Classifier (Perch)?

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

How accurate are these Bird Vocalization Classifier (Perch) speed estimates?

These are estimates with real error bars. The fastest result here, 260,633–695,023 tok/s on the B200, could reasonably land anywhere in its published range depending on which runtime you use.

What GPU do I need to run Bird Vocalization Classifier (Perch)?

The smallest card in our catalogue that holds Bird Vocalization Classifier (Perch) is the Tesla C1080, with 4 GB of memory. It runs the model at Q8_0 using about 0.7 GB, and produces roughly 4,726 tokens per second. 818 cards in total can run it.

How fast is Bird Vocalization Classifier (Perch) on a GPU?

It depends on the card. The quickest we calculate is a B200 at about 434,389 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 Bird Vocalization Classifier (Perch) clear that.

How much VRAM does Bird Vocalization Classifier (Perch) need?

About 0.7 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 Bird Vocalization Classifier (Perch) on a 8 GB GPU?

Yes. A CMP 170HX 8 GB with 8 GB runs it at Q8_0, using about 0.7 GB and generating roughly 80,905 tokens per second — a comfortable fit.

Source

Original publication

Record last updated 11 February 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