Calculate the TPS of the GeForce RTX 5090 Mobile on local AI models

NVIDIA 24 GB GDDR7 896 GB/s March 2025

Every model in our catalogue assessed against this card at the context length and minimum quality you choose. Speed is an estimate for a single request, calculated from this card's memory bandwidth and the size of each model once compressed.

Calculated for this card

532 models it can run

721 models in our catalogue altogether

Largest model it holds

Mixtral 8x7B

46.7B · Q3_K_M · 79.4 tok/s

Fastest model

Gemma 3 QAT 1B

379 tok/s · 1B

Which AI models can run on a GeForce RTX 5090 Mobile?

Set the inputs, read the answer

More context means more memory for the conversation cache. Speed is for a fresh conversation and does not change with this setting.

Hides models that would only fit by being compressed below this point.

532 models match

Calculating
Quantisation Fit
379 tok/s

323–455

Gemma 3 1B 1B Mar 2025 1.8 GB 33k tokens Q8_0 Comfortable
379 tok/s

323–455

Gemma 3 QAT 1B 1B Apr 2025 1.8 GB 33k tokens Q8_0 Comfortable
379 tok/s

228–607 · low confidence

HGRN 1B (WT 103) 1B Nov 2023 1.8 GB 131k tokens ? Q8_0 Comfortable
379 tok/s

228–607 · low confidence

LLama 3..2 Typhoon 2 1B 1B Dec 2024 1.8 GB 131k tokens ? Q8_0 Comfortable
379 tok/s

228–607 · low confidence

OLMo-1B 1B Feb 2024 1.8 GB 131k tokens ? Q8_0 Comfortable
379 tok/s

228–607 · low confidence

Pythia-1b 1B Apr 2023 1.8 GB 131k tokens ? Q8_0 Comfortable
351 tok/s

211–562 · low confidence

OpenELM-1.1B 1.1B May 2024 1.9 GB 131k tokens ? Q8_0 Comfortable
345 tok/s

207–552 · low confidence

DeciCoder-1B 1.1B Aug 2023 1.9 GB 131k tokens ? Q8_0 Comfortable
345 tok/s

207–552 · low confidence

SantaCoder 1.1B Jan 2023 1.9 GB 131k tokens ? Q8_0 Comfortable
345 tok/s

207–552 · low confidence

TinyLlama-1.1B (1T token checkpoint) 1.1B Oct 2023 1.9 GB 131k tokens ? Q8_0 Comfortable
345 tok/s

207–552 · low confidence

TinyLlama-1.1B (3T token checkpoint) 1.1B Oct 2023 1.9 GB 131k tokens ? Q8_0 Comfortable
316 tok/s

190–506 · low confidence

EXAONE 4.0 (1.2B) 1.2B Jul 2025 2.0 GB 131k tokens ? Q8_0 Comfortable
316 tok/s

190–506 · low confidence

LFM2-1.2B 1.2B Jul 2025 2.0 GB 131k tokens ? Q8_0 Comfortable
316 tok/s

190–506 · low confidence

MinerU2.5 1.2B Sep 2025 2.0 GB 131k tokens ? Q8_0 Comfortable
316 tok/s

190–506 · low confidence

Pleias 1.0 1.2B 1.2B Dec 2024 2.0 GB 131k tokens ? Q8_0 Comfortable
316 tok/s

190–506 · low confidence

Pleias-RAG-1B 1.2B Apr 2025 2.0 GB 131k tokens ? Q8_0 Comfortable
309 tok/s

262–370

Llama 3.2 1B 1.2B Sep 2024 2.2 GB 131k tokens Q8_0 Comfortable
304 tok/s

183–487 · low confidence

MiniCPM-1.2B 1.2B Jun 2024 2.0 GB 131k tokens ? Q8_0 Comfortable
292 tok/s

175–467 · low confidence

DeepSeek Coder 1.3B 1.3B Jan 2024 2.1 GB 131k tokens ? Q8_0 Comfortable
292 tok/s

175–467 · low confidence

DeepSeek-VL-1.3B 1.3B Mar 2024 2.1 GB 131k tokens ? Q8_0 Comfortable
292 tok/s

175–467 · low confidence

DigiRL 1.3B Jun 2024 2.1 GB 131k tokens ? Q8_0 Comfortable
292 tok/s

175–467 · low confidence

GLA Transformer 1.3B 1.3B Aug 2024 2.1 GB 131k tokens ? Q8_0 Comfortable
292 tok/s

175–467 · low confidence

Janus 1.3B 1.3B Oct 2024 2.1 GB 131k tokens ? Q8_0 Comfortable
292 tok/s

175–467 · low confidence

Kosmos-2.5 1.3B Aug 2024 2.1 GB 131k tokens ? Q8_0 Comfortable
292 tok/s

175–467 · low confidence

Otter 1.3B May 2023 2.1 GB 131k tokens ? 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

GeForce RTX 5090 Mobile full specification

Everything on record for this board, ordered by how much it bears on running a language model rather than by how a spec sheet would list it. Memory comes first because it decides the outcome; the rest is context.

Memory

The two specifications that decide what this card can run and how quickly. Capacity sets which models fit; bandwidth sets how many tokens per second they produce once they do.

Memory size
24 GB
Memory bandwidth
896 GB/s
Memory type
GDDR7
Memory bus width
256 bit
Memory clock
1.75 GHz

The chip

Which processor is on the board and how it was manufactured. A smaller process size generally means more performance for the same power.

Graphics processor
GB203
Architecture
Blackwell 2.0
Generation
GeForce 50 Mobile
Foundry
TSMC
Process size
5 nm
Transistors
45.6 billion
Transistor density
120,600 K/mm²
Die size
378 mm²
Released
27 March 2025

Clock speeds

How fast the processor runs. Worth far less here than on a gaming benchmark: generating text is limited by memory bandwidth, so a higher clock barely moves the result.

Base clock
990 MHz
Boost clock
1.52 GHz

Processing units

What the chip contains. These drive graphics performance and matter mainly for processing a long prompt rather than for producing the answer.

Shading units
10,496
Texture mapping units
328
Render output units
112
Streaming multiprocessors
82
Tensor cores
328
Ray tracing cores
82
L1 cache
128 KB
L2 cache
64 MB

Theoretical performance

Peak arithmetic rates published for the board. These are ceilings that no real workload reaches, and generating text reaches a small fraction of them because it is limited by memory rather than arithmetic.

Half precision (FP16)
31.8 TFLOPS
Single precision (FP32)
31.8 TFLOPS
Double precision (FP64)
496.9 GFLOPS
Pixel rate
170 GPixel/s
Texture rate
497 GTexel/s

The board

What it takes to physically install and power the card — the practical constraints that decide whether it fits the machine you already own.

Power draw (TDP)
95 W
Power connectors
None
Bus interface
PCIe 5.0 x16
Slot width
IGP

Software support

Which graphics and compute interfaces the card supports. CUDA compute capability is the one that bears on inference: below 7.0 there are no tensor cores, and modern inference software falls back to slower code paths.

CUDA compute capability
12.0
DirectX
12.2
OpenGL
4.6
Vulkan
1.4
OpenCL
3.0
Shader model
6.8

Listings

Where to buy a GeForce RTX 5090 Mobile

No vendor is currently listing this card. Listings come from vendors who publish them here directly — browse the vendor directory to see who is selling what.

What the numbers mean

What the memory subsystem means for AI

Memory

24 GB

Bandwidth

896 GB/s

Largest model

Mixtral 8x7B

GeForce RTX 5090 Mobile carries 24 GB of GDDR7. That covers the mid-sized models most people actually run. Once the runtime and driver reserve their working space, roughly this much is left: 21.6 GB.

Memory bandwidth reaches 896 GB/s across a bus of 256 bits. Generating a token means reading every weight once, so that figure sets the pace more than any other number here, and at this level text arrives faster than most people read.

The figure is the bus width multiplied by a memory clock of 1.75 GHz. Both halves matter, and neither is visible in a gaming benchmark.

The practical ceiling is Mixtral 8x7B, 46.7B, compressed to Q3_K_M and generating around 79.4 tokens per second.

The chip and how it was built

GeForce RTX 5090 Mobile is built on the graphics processor GB203, using the architecture Blackwell 2.0 from NVIDIA, as part of the generation GeForce 50 Mobile.

The chip is manufactured by TSMC, on a process of 5 nm, with a die measuring 378 mm², holding 45.6 billion transistors. A smaller process generally means more performance for the same power, though for language models it matters far less than the memory subsystem.

It was released in March 2025, roughly 1.4672998138926 years ago. Inference software support tends to follow hardware by a year or two, so a card of this age generally has mature, well-optimised code paths available to it.

Compute throughput, and why it matters less than it looks

FP16

31.8 TFLOPS

FP64

496.9 GFLOPS

Tensor cores

328

On paper GeForce RTX 5090 Mobile reaches 31.8 TFLOPS at half precision, and 31.8 TFLOPS at single precision. These are peak figures no real workload sustains, and generating text reaches only a small fraction of them — decoding is limited by memory rather than arithmetic, which is why a card can look enormously powerful here and still produce tokens at an ordinary rate.

Double-precision throughput reaches 496.9 GFLOPS. It has no bearing on running a language model — no inference runtime uses it — but it separates datacentre parts from consumer ones, since the latter deliberately restrict it.

The card carries 328 tensor cores across 82 streaming multiprocessors. These accelerate the matrix arithmetic at the heart of a transformer, and they are what make prompt processing — reading a long document before answering — dramatically faster than it would otherwise be.

Clocks run from a base of 990 MHz to a boost of 1.52 GHz. Worth far less here than on a gaming benchmark: raising the clock speeds up the arithmetic, and the arithmetic is not what generation is waiting on.

Cache and processing units

GeForce RTX 5090 Mobile has an L1 cache of 128 KB, backed by an L2 cache of 64 MB. Cache absorbs a share of the memory traffic that would otherwise hit the main bus, which is the one place on this page where a number other than bandwidth quietly affects generation speed — a large L2 lets more of the working set stay close to the cores.

There are 10,496 shading units, 328 texture mapping units, and 112 render output units. These drive graphics workloads and contribute to prompt processing, but they sit idle for much of the time a model spends generating a reply.

Power, size and installation

Power draw

95 W

GeForce RTX 5090 Mobile is rated at 95 W. Running a language model keeps a card busy in bursts rather than continuously — it draws hard while generating and idles between requests — so sustained draw over a working day is usually well below the rated figure.

The board occupies igp. Worth checking against the case and power supply already in the machine, since the largest cards need considerably more of both than a typical desktop provides.

It connects over PCIe 5.0 x16. The interface governs how quickly a model is loaded from disk into the card, not how fast it runs once there, so a narrower link costs a few seconds at startup and nothing thereafter.

The extremes

The largest AI models that run on a GeForce RTX 5090 Mobile

The biggest open-weight models that fit on this card, newest first. Each is shown at the best compression the card can hold.

  1. 01 Qwen3.6-35B-A3B 35B · Q4_K_M · Apr 2026 139 tok/s
  2. 02 Qwen3-Omni-30B-A3B 35.3B · Q4_K_M · Sep 2025 138 tok/s
  3. 03 Seed-OSS-36B-Base 36B · IQ4_XS · Aug 2025 25.9 tok/s
  4. 04 InternVL2_5-38B 38.4B · Q3_K_M · Dec 2024 26.7 tok/s
  5. 05 InternVL2-40B 40.1B · Q3_K_M · Jul 2024 25.5 tok/s
  6. 06 JIUTIAN-139MoE 38.8B · Q3_K_M · Jun 2024 26.4 tok/s
  7. 07 VILA1.5-40B 40B · Q3_K_M · May 2024 25.6 tok/s
  8. 08 Mixtral 8x7B 46.7B · Q3_K_M · Dec 2023 79.4 tok/s
  9. 09 Falcon-40B 40B · Q3_K_M · Mar 2023 25.6 tok/s
  10. 10 gpt-sw3-40b 40B · Q3_K_M · Mar 2023 25.6 tok/s

The fastest AI models on a GeForce RTX 5090 Mobile

Where this card produces tokens quickest. Smaller models dominate here, because generating each token means reading the whole model out of memory once.

  1. 01 Gemma 3 QAT 1B 1B · Q8_0 · 1.8 GB 379 tok/s
  2. 02 Gemma 3 1B 1B · Q8_0 · 1.8 GB 379 tok/s
  3. 03 LLama 3..2 Typhoon 2 1B 1B · Q8_0 · 1.8 GB 379 tok/s
  4. 04 OLMo-1B 1B · Q8_0 · 1.8 GB 379 tok/s
  5. 05 HGRN 1B (WT 103) 1B · Q8_0 · 1.8 GB 379 tok/s
  6. 06 Pythia-1b 1B · Q8_0 · 1.8 GB 379 tok/s
  7. 07 OpenELM-1.1B 1.1B · Q8_0 · 1.9 GB 351 tok/s
  8. 08 TinyLlama-1.1B (1T token checkpoint) 1.1B · Q8_0 · 1.9 GB 345 tok/s
  9. 09 TinyLlama-1.1B (3T token checkpoint) 1.1B · Q8_0 · 1.9 GB 345 tok/s
  10. 10 DeciCoder-1B 1.1B · Q8_0 · 1.9 GB 345 tok/s

Step by step

How to work out the tokens per second of a GeForce RTX 5090 Mobile

You do not have to calculate anything by hand — the gputps.com calculator on this page has already worked it out for every model this card can hold. Reading off the answer takes six steps.

  1. 01

    Start with the model, not the specification

    The table lists 532 models the card handles. The search box takes a name or a size such as 27b, which matches on parameter count.

  2. 02

    Decide how long your conversations run

    Set the context to your real working length. Short questions cost almost nothing, but a long document can consume a large share of 24 GB that is frequently the difference between a model fitting and not.

  3. 03

    Set a minimum quality if you need one

    Each model is shown at the best compression this card can hold. A minimum quality hides the ones that only fit by being squeezed further than you would accept.

  4. 04

    Read the speed and the range

    Speeds come with error bars for a reason. The best case here is 379 tok/s on Gemma 3 QAT 1B. The same card and model vary by thirty to fifty per cent between inference runtimes.

  5. 05

    Check the headroom before you decide

    A tight fit runs but leaves no room to raise the context later; comfortable has headroom. The memory column shows what each model needs against an available 24 GB.

  6. 06

    Cross-check against other hardware

    Following a model through to its own page lists all the hardware that can run it, so you can see how it compares against GeForce RTX 5090 Mobile.

Answers

GeForce RTX 5090 Mobile — common questions

01

GeForce RTX 5090 Mobile— what is the largest AI model it can run?

The largest model in our catalogue that fits is Mixtral 8x7B at 46.7B parameters, compressed to Q3_K_M. It generates roughly 79.4 tokens per second and needs about 21.0 GB of the card's memory.

02

GeForce RTX 5090 Mobile— how many tokens per second does it produce?

It depends on the model. The fastest model we track here is Gemma 3 QAT 1B at about 379 tokens per second, while larger models run proportionally slower because each token requires reading the whole model out of memory once. Speeds are estimates for a single conversation at a time.

03

GeForce RTX 5090 Mobile— can it run 7B models?

Yes. For example it runs Gemma 4 E4B at Q8_0, using about 9.8 GB of memory and generating around 84.3 tokens per second.

04

GeForce RTX 5090 Mobile— can it run 13B models?

Yes. For example it runs DeepSeekMoE-16B at Q8_0, using about 17.5 GB of memory and generating around 132 tokens per second.

05

GeForce RTX 5090 Mobile— can it run 30B models?

Yes. For example it runs Nemotron 3-Nano-30B-A3B at Q4_K_M, using about 18.1 GB of memory and generating around 154 tokens per second.

06

GeForce RTX 5090 Mobile— how much memory does it have?

This card has 24 GB of GDDR7. Around a tenth is reserved by the inference runtime and the driver, leaving roughly 21.6 GB available for a model and its conversation.

07

GeForce RTX 5090 Mobile— what is its memory bandwidth?

Memory bandwidth reaches 896 GB/s across a bus of 256 bits. This is the single best predictor of how fast it generates text, because producing each token means reading the entire model out of memory once.

08

GeForce RTX 5090 Mobile— what type of memory does it use?

It uses GDDR7 clocked at 1.75 GHz. HBM types are found on datacentre accelerators and carry far more bandwidth than the GDDR used on desktop cards, which is why they generate tokens considerably faster at the same capacity.

09

GeForce RTX 5090 Mobile— who makes it?

This is a product of NVIDIA, with the chip manufactured by TSMC, on a process of 5 nm.

10

GeForce RTX 5090 Mobile— when was it released?

It was released in March 2025.

11

GeForce RTX 5090 Mobile— how much power does it use?

Rated board power is 95 W. Generating text draws hard in bursts and idles between requests, so average consumption over a working session is normally well below the rated figure.

12

GeForce RTX 5090 Mobile— how much cache does it have?

The L1 cache is 128 KB, and the L2 cache is 64 MB. Cache absorbs part of the memory traffic that would otherwise reach the main bus, so a larger L2 gives a modest lift to generation speed beyond what bandwidth alone predicts.

13

GeForce RTX 5090 Mobile— what are its TFLOPS?

It is rated at 31.8 TFLOPS at half precision and 31.8 TFLOPS at single precision. These are peak arithmetic ceilings rather than achievable rates, and text generation reaches only a small fraction of them because it is limited by memory bandwidth instead.

14

GeForce RTX 5090 Mobile— how many tensor cores does it have?

It has 328 tensor cores across 82 streaming multiprocessors. They accelerate the matrix arithmetic a transformer is built from, which mainly speeds up processing a long prompt rather than producing the reply.

15

GeForce RTX 5090 Mobile— does it support CUDA?

Yes. It reports CUDA compute capability 12.0. Capability 7.0 and above has tensor cores, which modern inference software uses; below that it falls back to slower code paths for quantised models.

16

GeForce RTX 5090 Mobile— what bus interface does it use?

It uses PCIe 5.0 x16. This governs how fast a model is loaded onto the card rather than how fast it runs once loaded, so it costs a few seconds at startup and nothing during generation.

17

GeForce RTX 5090 Mobile— is it good for running local AI models?

Its memory comfortably covers the mid-sized models most people run locally and its bandwidth is high enough to generate text faster than most people read. In total it runs 532 of the models we track. Whether that is enough depends entirely on which model you want — the table above answers that directly.

18

GeForce RTX 5090 Mobile— can it run a model that does not fit in its memory?

It can be split, with the overflow held in system memory beyond the card's 24 GB drags the whole thing down, and none of the figures on this page assume it.

19

Would two GeForce RTX 5090 Mobile cards be twice as fast?

Capacity adds, throughput does not. Two of them give you 48 GB which lets you hold models neither could hold alone, but generation does not split that way. These figures describe one card.

20

GeForce RTX 5090 Mobile— which AI models can it run?

532 of the 721 open-weight language models we track fit on this card and can be run locally. The table on this page lists every one, with the memory it needs, the quantisation it runs at and an estimated generation speed.

The other direction

Looking at it from the other side?

This page starts from the hardware. If you already know which model you want and need to know what it takes to run it, start from the model instead.

All GPUs