Sailor-7B-Chat TPS calculator

Open weights Sea AI Lab,Singapore University of Technology & Design 7.7B parameters April 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

589 cards that can run it

818 cards we hold specifications for

Smallest card that fits

Tesla K20c

5 GB · Q3_K_M · 26.2 tok/s

Fastest card

B200

439 tok/s · 180 GB

Which GPUs can run Sailor-7B-Chat?

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

589 cards match

Calculating

					Needs	Quantisation	Fit
439 tok/s 263–702 · low confidence	B200 NVIDIA	180 GB	8,000 GB/s	Jan 2024	9.0 GB	Q8_0	Comfortable
439 tok/s 263–702 · low confidence	B300 NVIDIA	288 GB	8,000 GB/s	Sep 2025	9.0 GB	Q8_0	Comfortable
350 tok/s 210–561 · low confidence	Radeon Instinct MI350X AMD	288 GB	8,190 GB/s	Jan 2025	9.0 GB	Q8_0	Comfortable
350 tok/s 210–561 · low confidence	Radeon Instinct MI355X AMD	288 GB	8,190 GB/s	Jan 2025	9.0 GB	Q8_0	Comfortable
280 tok/s 168–448 · low confidence	Radeon Instinct MI300 AMD	128 GB	6,550 GB/s	Jan 2023	9.0 GB	Q8_0	Comfortable
268 tok/s 161–429 · low confidence	H200 NVL NVIDIA	141 GB	4,890 GB/s	Nov 2024	9.0 GB	Q8_0	Comfortable
268 tok/s 161–429 · low confidence	H200 SXM 141 GB NVIDIA	141 GB	4,890 GB/s	Nov 2024	9.0 GB	Q8_0	Comfortable
257 tok/s 154–411 · low confidence	Radeon Instinct MI325X AMD	256 GB	6,000 GB/s	Oct 2024	9.0 GB	Q8_0	Comfortable
228 tok/s 137–365 · low confidence	Radeon Instinct MI300A AMD	128 GB	5,325 GB/s	Dec 2023	9.0 GB	Q8_0	Comfortable
228 tok/s 137–365 · low confidence	Radeon Instinct MI300X AMD	192 GB	5,325 GB/s	Dec 2023	9.0 GB	Q8_0	Comfortable
228 tok/s 137–365 · low confidence	Radeon Instinct MI308X AMD	192 GB	5,325 GB/s	Dec 2023	9.0 GB	Q8_0	Comfortable
216 tok/s 130–346 · low confidence	H100 NVL 94 GB NVIDIA	94 GB	3,940 GB/s	Mar 2023	9.0 GB	Q8_0	Comfortable
184 tok/s 111–295 · low confidence	H100 PCIe 96 GB NVIDIA	96 GB	3,360 GB/s	Mar 2023	9.0 GB	Q8_0	Comfortable
184 tok/s 111–295 · low confidence	H100 SXM5 80 GB NVIDIA	80 GB	3,360 GB/s	Oct 2022	9.0 GB	Q8_0	Comfortable
184 tok/s 111–295 · low confidence	H100 SXM5 94 GB NVIDIA	94 GB	3,360 GB/s	Mar 2023	9.0 GB	Q8_0	Comfortable
184 tok/s 111–295 · low confidence	H100 SXM5 96 GB NVIDIA	96 GB	3,360 GB/s	Mar 2023	9.0 GB	Q8_0	Comfortable
184 tok/s 111–295 · low confidence	H800 SXM5 NVIDIA	80 GB	3,360 GB/s	Mar 2023	9.0 GB	Q8_0	Comfortable
140 tok/s 84–225 · low confidence	Radeon Instinct MI250 AMD	128 GB	3,280 GB/s	Nov 2021	9.0 GB	Q8_0	Comfortable
140 tok/s 84–225 · low confidence	Radeon Instinct MI250X AMD	128 GB	3,280 GB/s	Nov 2021	9.0 GB	Q8_0	Comfortable
119 tok/s 71–190 · low confidence	CMP 170HX 8 GB NVIDIA	8 GB	1,490 GB/s	Sep 2021	7.2 GB	Q6_K	Tight
117 tok/s 70–187 · low confidence	Data Center GPU Max 1550 Intel	128 GB	3,280 GB/s	Jan 2023	9.0 GB	Q8_0	Comfortable
114 tok/s 69–183 · low confidence	Data Center GPU Max Subsystem Intel	128 GB	3,210 GB/s	Jan 2023	9.0 GB	Q8_0	Comfortable
112 tok/s 67–179 · low confidence	A100 SXM4 80 GB NVIDIA	80 GB	2,040 GB/s	Nov 2020	9.0 GB	Q8_0	Comfortable
112 tok/s 67–179 · low confidence	A100X NVIDIA	80 GB	2,040 GB/s	Jun 2021	9.0 GB	Q8_0	Comfortable
112 tok/s 67–179 · low confidence	A800 SXM4 80 GB NVIDIA	80 GB	2,040 GB/s	Aug 2022	9.0 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: Sea AI Lab,Singapore University of Technology & Design
Organisation type: Academia
Country: Singapore
Published: 4 April 2024
Authors: Longxu Dou, Qian Liu, Guangtao Zeng, Jia Guo, Jiahui Zhou, Wei Lu, Min Lin

What it does

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

Domain: Language
Task: Chat
Base model: Qwen1.5-7B

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: 7.7B
Training data: tokens
Epochs: 1
Batch size: 4,000,000

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: 1.8 × 10²³ FLOP

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 40 GB
Chips used: 64
Power draw: 50.6 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 (unrestricted)
Training code: Unreleased
Hugging Face: sail

How it is classified

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

Likely above 10²³ FLOP: Yes
Record confidence: Confident

Sources

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

Reference: Sailor: Open Language Models for South-East Asia
Last updated: 28 November 2025

The extremes

The ten fastest GPUs that run Sailor-7B-Chat

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 Sailor-7B-Chat

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

The hardware side

Minimum card

Tesla K20c

Memory needed

4.5 GB

Fastest

439 tok/s

Sailor-7B-Chat is small enough at 7.7B parameters that hardware is rarely the obstacle — 589 of the cards we track can run it, including cards several years old.

At the low end, a Tesla K20c handles it — 5 GB, at Q3_K_M, for about 26.2 tokens per second.

The quickest result comes from a B200 at around 439 tokens per second — its 8,000 GB/s of bandwidth is what buys that.

Background

Sailor-7B-Chat was published by Sea AI Lab,Singapore University of Technology & Design, in Singapore, in April 2024. The organisation is categorised as academia.

It works in Language, and is recorded as doing chat.

Its starting point was Qwen1.5-7B — 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. It is published under the sail organisation on Hugging Face.

Reading the throughput figures

The median result is around 23.7 tokens per second; 546 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.

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 1.8 × 10²³ FLOP of computation, on NVIDIA A100 SXM4 40 GB — a measure of what producing the model cost, not of how fast it answers.

Step by step

How to choose a GPU for Sailor-7B-Chat

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

Every card here has been checked against Sailor-7B-Chat — around 4.5 GB at Q3_K_M. Capacity is the gate — a card either holds it or it does not.
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 Sailor-7B-Chat.
03

Set a quality floor

The quantisation column varies by card, because a bigger card holds a more accurate copy of Sailor-7B-Chat — Q3_K_M 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 Sailor-7B-Chat is effectively an ordering by memory bandwidth, which is why the B200 tops it at 439 tok/s.
05

Read the fit column last

Tight means Sailor-7B-Chat 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

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 Sailor-7B-Chat alone — a card is usually bought for more than one model.

Answers

Sailor-7B-Chat — common questions

Where can I download Sailor-7B-Chat?

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

How much compute was used to train Sailor-7B-Chat?

Around 1.8 × 10²³ FLOP, on NVIDIA A100 SXM4 40 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.

Can I run Sailor-7B-Chat if it does not fit in my GPU?

It can be split between the card and system memory, but Sailor-7B-Chat generates painfully slowly that way — the nearest miss we calculate is short by 1.8 GB. Nothing on this page assumes offloading.

Would two GPUs run Sailor-7B-Chat faster?

A second card roughly doubles the memory available but not the generation rate. With 589 cards already able to run Sailor-7B-Chat alone, the case for pairing is weak.

Why does the quantisation differ between cards for Sailor-7B-Chat?

Because capacity varies, so does how hard Sailor-7B-Chat has to be squeezed — 4 distinct levels appear in the table above. Set a minimum quality to compare at one.

How accurate are these Sailor-7B-Chat 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 263–702 tok/s on the B200 rather than a single number.

What GPU do I need to run Sailor-7B-Chat?

The smallest card in our catalogue that holds Sailor-7B-Chat is the Tesla K20c, with 5 GB of memory. It runs the model at Q3_K_M using about 4.5 GB, and produces roughly 26.2 tokens per second. 589 cards in total can run it.

How fast is Sailor-7B-Chat on a GPU?

It depends on the card. The quickest we calculate is a B200 at about 439 tokens per second; the slowest that still runs it manages considerably less. Reading speed is around ten tokens per second, and 546 of the cards that can run Sailor-7B-Chat clear that.

How much VRAM does Sailor-7B-Chat need?

About 4.5 GB at Q3_K_M 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 Sailor-7B-Chat on a 8 GB GPU?

Yes. A CMP 170HX 8 GB with 8 GB runs it at Q6_K, using about 7.2 GB and generating roughly 119 tokens per second — a tight fit.

Can I run Sailor-7B-Chat on a 12 GB GPU?

Yes. A GeForce RTX 3080 Ti with 12 GB runs it at Q8_0, using about 9.0 GB and generating roughly 50.1 tokens per second — a comfortable fit.

Can I run Sailor-7B-Chat on a 16 GB GPU?

Yes. A Tesla V100 SXM2 16 GB with 16 GB runs it at Q8_0, using about 9.0 GB and generating roughly 62.0 tokens per second — a comfortable fit.

Can I run Sailor-7B-Chat on a 24 GB GPU?

Yes. A GeForce RTX 5090 D V2 with 24 GB runs it at Q8_0, using about 9.0 GB and generating roughly 73.5 tokens per second — a comfortable fit.

Is Sailor-7B-Chat open source?

Its weights are published, so Sailor-7B-Chat 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 Sailor-7B-Chat have?

Sailor-7B-Chat has 7.7B parameters. https://huggingface.co/sail/Sailor-7B-Chat. 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 Sailor-7B-Chat?

Sailor-7B-Chat was published by Sea AI Lab,Singapore University of Technology & Design, based in Singapore, categorised as academia.

When was Sailor-7B-Chat released?

Sailor-7B-Chat was published in April 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.

What is Sailor-7B-Chat used for?

Sailor-7B-Chat works in Language, and is recorded as handling chat. These are the areas it was designed around; they describe intent rather than a hard boundary.

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