Refact-1.6B TPS calculator

Open weights Refact AI 1.6B parameters August 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 · 23.0 tok/s

Fastest card

B200

2,118 tok/s · 180 GB

Which GPUs can run Refact-1.6B?

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
2,118 tok/s 1,271–3,388 · low confidence	B200 NVIDIA	180 GB	8,000 GB/s	Jan 2024	2.4 GB	Q8_0	Comfortable
2,118 tok/s 1,271–3,388 · low confidence	B300 NVIDIA	288 GB	8,000 GB/s	Sep 2025	2.4 GB	Q8_0	Comfortable
1,691 tok/s 1,015–2,706 · low confidence	Radeon Instinct MI350X AMD	288 GB	8,190 GB/s	Jan 2025	2.4 GB	Q8_0	Comfortable
1,691 tok/s 1,015–2,706 · low confidence	Radeon Instinct MI355X AMD	288 GB	8,190 GB/s	Jan 2025	2.4 GB	Q8_0	Comfortable
1,352 tok/s 811–2,164 · low confidence	Radeon Instinct MI300 AMD	128 GB	6,550 GB/s	Jan 2023	2.4 GB	Q8_0	Comfortable
1,294 tok/s 777–2,071 · low confidence	H200 NVL NVIDIA	141 GB	4,890 GB/s	Nov 2024	2.4 GB	Q8_0	Comfortable
1,294 tok/s 777–2,071 · low confidence	H200 SXM 141 GB NVIDIA	141 GB	4,890 GB/s	Nov 2024	2.4 GB	Q8_0	Comfortable
1,239 tok/s 743–1,982 · low confidence	Radeon Instinct MI325X AMD	256 GB	6,000 GB/s	Oct 2024	2.4 GB	Q8_0	Comfortable
1,099 tok/s 660–1,759 · low confidence	Radeon Instinct MI300A AMD	128 GB	5,325 GB/s	Dec 2023	2.4 GB	Q8_0	Comfortable
1,099 tok/s 660–1,759 · low confidence	Radeon Instinct MI300X AMD	192 GB	5,325 GB/s	Dec 2023	2.4 GB	Q8_0	Comfortable
1,099 tok/s 660–1,759 · low confidence	Radeon Instinct MI308X AMD	192 GB	5,325 GB/s	Dec 2023	2.4 GB	Q8_0	Comfortable
1,043 tok/s 626–1,669 · low confidence	H100 NVL 94 GB NVIDIA	94 GB	3,940 GB/s	Mar 2023	2.4 GB	Q8_0	Comfortable
889 tok/s 534–1,423 · low confidence	H100 PCIe 96 GB NVIDIA	96 GB	3,360 GB/s	Mar 2023	2.4 GB	Q8_0	Comfortable
889 tok/s 534–1,423 · low confidence	H100 SXM5 80 GB NVIDIA	80 GB	3,360 GB/s	Oct 2022	2.4 GB	Q8_0	Comfortable
889 tok/s 534–1,423 · low confidence	H100 SXM5 94 GB NVIDIA	94 GB	3,360 GB/s	Mar 2023	2.4 GB	Q8_0	Comfortable
889 tok/s 534–1,423 · low confidence	H100 SXM5 96 GB NVIDIA	96 GB	3,360 GB/s	Mar 2023	2.4 GB	Q8_0	Comfortable
889 tok/s 534–1,423 · low confidence	H800 SXM5 NVIDIA	80 GB	3,360 GB/s	Mar 2023	2.4 GB	Q8_0	Comfortable
677 tok/s 406–1,084 · low confidence	Radeon Instinct MI250 AMD	128 GB	3,280 GB/s	Nov 2021	2.4 GB	Q8_0	Comfortable
677 tok/s 406–1,084 · low confidence	Radeon Instinct MI250X AMD	128 GB	3,280 GB/s	Nov 2021	2.4 GB	Q8_0	Comfortable
564 tok/s 339–903 · low confidence	Data Center GPU Max 1550 Intel	128 GB	3,280 GB/s	Jan 2023	2.4 GB	Q8_0	Comfortable
552 tok/s 331–884 · low confidence	Data Center GPU Max Subsystem Intel	128 GB	3,210 GB/s	Jan 2023	2.4 GB	Q8_0	Comfortable
540 tok/s 324–864 · low confidence	A100 SXM4 80 GB NVIDIA	80 GB	2,040 GB/s	Nov 2020	2.4 GB	Q8_0	Comfortable
540 tok/s 324–864 · low confidence	A100X NVIDIA	80 GB	2,040 GB/s	Jun 2021	2.4 GB	Q8_0	Comfortable
540 tok/s 324–864 · low confidence	A800 SXM4 80 GB NVIDIA	80 GB	2,040 GB/s	Aug 2022	2.4 GB	Q8_0	Comfortable
540 tok/s 324–864 · 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: Refact AI
Organisation type: Industry
Country: United Kingdom of Great Britain and Northern Ireland
Published: 29 August 2023
Authors: Oleg Klimov, Sergey Vakhreev

What it does

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

Domain: Language
Task: Language modeling/generation
Approach: Self-supervised learning

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: 1,200,000,000,000 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: 1.2 × 10²² FLOP
How it was established: Operation counting,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 RTX A5000
Chips used: 6
Chip-hours: 4,032
Wall-clock time: 672 hours (28 days)
Power draw: 2.7 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 (restricted use)
Training code: Unreleased

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: Refact-1.6B
Last updated: 28 November 2025

The extremes

The ten fastest GPUs that run Refact-1.6B

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 Refact-1.6B

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 it takes to run this model

Minimum card

Tesla C1080

Memory needed

2.4 GB

Fastest

2,118 tok/s

Refact-1.6B 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 smallest card that holds it is the Tesla C1080 with 4 GB, running it at Q8_0 and producing around 23.0 tokens per second.

The quickest result comes from a B200 at around 2,118 tokens per second — its 8,000 GB/s of bandwidth is what buys that.

About this model

Refact-1.6B was published by Refact AI, in United Kingdom of Great Britain and Northern Ireland, in August 2023. The organisation is categorised as industry.

It works in Language, and is recorded as doing language modeling/generation.

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.

How fast it runs, and why

Half the cards that hold it manage more than 59.5 tokens per second, and 794 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.

Memory here is estimated from size rather than computed from the architecture, which is not recorded for this model — the numbers are indicative rather than exact.

What went into building it

Producing it required around 1.2 × 10²² FLOP of arithmetic, on NVIDIA RTX A5000, which is a statement about the training budget rather than about inference.

The training set ran to roughly 1,200,000,000,000 tokens.

Step by step

How to choose a GPU for Refact-1.6B

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 Refact-1.6B — around 2.4 GB at Q8_0. Capacity is the gate — a card either holds it or it does not.
02

Match the context to your actual use

The conversation occupies memory too, and grows as it goes. Set the slider to the length you expect: at long context Refact-1.6B 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 Refact-1.6B — Q8_0 on the smallest card that fits. Set a floor to hold the comparison at one level.
04

Rank by throughput rather than spec sheet

The speed ordering for Refact-1.6B is effectively an ordering by memory bandwidth, which is why the B200 tops it at 2,118 tok/s.
05

Check the fit verdict before buying

Tight means Refact-1.6B loads and works, with no room to raise the context later. Comfortable means you can. The difference matters more than a few tokens per second.
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 Refact-1.6B is settled.

Answers

Refact-1.6B — common questions

When was Refact-1.6B released?

Refact-1.6B was published in August 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 Refact-1.6B used for?

Refact-1.6B works in Language, and is recorded as handling language modeling/generation. Models frequently carry more than one of each, and the tags describe purpose rather than capability limits.

Where can I download Refact-1.6B?

The weights for Refact-1.6B 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 Refact-1.6B?

Around 1.2 × 10²² FLOP, on NVIDIA RTX A5000. 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 Refact-1.6B if it does not fit in my GPU?

It can be split between the card and system memory, but Refact-1.6B generates painfully slowly that way. Nothing on this page assumes offloading.

Would two GPUs run Refact-1.6B faster?

Two cards buy memory rather than speed. That matters for Refact-1.6B only if one card cannot hold it — 818 can, so a second adds little.

Why does the quantisation differ between cards for Refact-1.6B?

Because capacity varies, so does how hard Refact-1.6B has to be squeezed — 1 distinct levels appear in the table above. Set a minimum quality to compare at one.

How accurate are these Refact-1.6B 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 1,271–3,388 tok/s on the B200 rather than a single number.

What GPU do I need to run Refact-1.6B?

The smallest card in our catalogue that holds Refact-1.6B is the Tesla C1080, with 4 GB of memory. It runs the model at Q8_0 using about 2.4 GB, and produces roughly 23.0 tokens per second. 818 cards in total can run it.

How fast is Refact-1.6B on a GPU?

It depends on the card. The quickest we calculate is a B200 at about 2,118 tokens per second; the slowest that still runs it manages considerably less. Reading speed is around ten tokens per second, and 794 of the cards that can run Refact-1.6B clear that.

How much VRAM does Refact-1.6B 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 Refact-1.6B 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 394 tokens per second — a comfortable fit.

Can I run Refact-1.6B 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 242 tokens per second — a comfortable fit.

Can I run Refact-1.6B 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 299 tokens per second — a comfortable fit.

Can I run Refact-1.6B 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 355 tokens per second — a comfortable fit.

Is Refact-1.6B open source?

Its weights are published, so Refact-1.6B 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 Refact-1.6B have?

Refact-1.6B has 1.6B parameters. 1.6B. 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 Refact-1.6B?

Refact-1.6B was published by Refact AI, based in United Kingdom of Great Britain and Northern Ireland, categorised as industry.

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