Cosmos-Predict1-14b-Video2World TPS calculator
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 we hold specifications for
Smallest card that fits
P102-101
10 GB · IQ4_XS · 20.2 tok/s
Fastest card
B200
242 tok/s · 180 GB
Which GPUs can run Cosmos-Predict1-14b-Video2World?
Set the inputs, read the answer
A longer conversation needs more memory, which can push this model off smaller cards.
Hides cards that would only fit the model by compressing it below this point.
306 cards match
Calculating| Needs | Quantisation | Fit | |||||
|---|---|---|---|---|---|---|---|
|
242
tok/s
145–387 · low confidence |
B200 NVIDIA | 180 GB | 8,000 GB/s | Jan 2024 | 15.7 GB | Q8_0 | Comfortable |
|
242
tok/s
145–387 · low confidence |
B300 NVIDIA | 288 GB | 8,000 GB/s | Sep 2025 | 15.7 GB | Q8_0 | Comfortable |
|
193
tok/s
116–309 · low confidence |
Radeon Instinct MI350X AMD | 288 GB | 8,190 GB/s | Jan 2025 | 15.7 GB | Q8_0 | Comfortable |
|
193
tok/s
116–309 · low confidence |
Radeon Instinct MI355X AMD | 288 GB | 8,190 GB/s | Jan 2025 | 15.7 GB | Q8_0 | Comfortable |
|
155
tok/s
93–247 · low confidence |
Radeon Instinct MI300 AMD | 128 GB | 6,550 GB/s | Jan 2023 | 15.7 GB | Q8_0 | Comfortable |
|
148
tok/s
89–237 · low confidence |
H200 NVL NVIDIA | 141 GB | 4,890 GB/s | Nov 2024 | 15.7 GB | Q8_0 | Comfortable |
|
148
tok/s
89–237 · low confidence |
H200 SXM 141 GB NVIDIA | 141 GB | 4,890 GB/s | Nov 2024 | 15.7 GB | Q8_0 | Comfortable |
|
142
tok/s
85–227 · low confidence |
Radeon Instinct MI325X AMD | 256 GB | 6,000 GB/s | Oct 2024 | 15.7 GB | Q8_0 | Comfortable |
|
126
tok/s
75–201 · low confidence |
Radeon Instinct MI300A AMD | 128 GB | 5,325 GB/s | Dec 2023 | 15.7 GB | Q8_0 | Comfortable |
|
126
tok/s
75–201 · low confidence |
Radeon Instinct MI300X AMD | 192 GB | 5,325 GB/s | Dec 2023 | 15.7 GB | Q8_0 | Comfortable |
|
126
tok/s
75–201 · low confidence |
Radeon Instinct MI308X AMD | 192 GB | 5,325 GB/s | Dec 2023 | 15.7 GB | Q8_0 | Comfortable |
|
119
tok/s
72–191 · low confidence |
H100 NVL 94 GB NVIDIA | 94 GB | 3,940 GB/s | Mar 2023 | 15.7 GB | Q8_0 | Comfortable |
|
116
tok/s
70–185 · low confidence |
CMP 170HX 10 GB NVIDIA | 10 GB | 1,560 GB/s | Sep 2021 | 8.4 GB | IQ4_XS | Tight |
|
102
tok/s
61–163 · low confidence |
H100 PCIe 96 GB NVIDIA | 96 GB | 3,360 GB/s | Mar 2023 | 15.7 GB | Q8_0 | Comfortable |
|
102
tok/s
61–163 · low confidence |
H100 SXM5 80 GB NVIDIA | 80 GB | 3,360 GB/s | Oct 2022 | 15.7 GB | Q8_0 | Comfortable |
|
102
tok/s
61–163 · low confidence |
H100 SXM5 94 GB NVIDIA | 94 GB | 3,360 GB/s | Mar 2023 | 15.7 GB | Q8_0 | Comfortable |
|
102
tok/s
61–163 · low confidence |
H100 SXM5 96 GB NVIDIA | 96 GB | 3,360 GB/s | Mar 2023 | 15.7 GB | Q8_0 | Comfortable |
|
102
tok/s
61–163 · low confidence |
H800 SXM5 NVIDIA | 80 GB | 3,360 GB/s | Mar 2023 | 15.7 GB | Q8_0 | Comfortable |
|
77.4
tok/s
46–124 · low confidence |
Radeon Instinct MI250 AMD | 128 GB | 3,280 GB/s | Nov 2021 | 15.7 GB | Q8_0 | Comfortable |
|
77.4
tok/s
46–124 · low confidence |
Radeon Instinct MI250X AMD | 128 GB | 3,280 GB/s | Nov 2021 | 15.7 GB | Q8_0 | Comfortable |
|
64.5
tok/s
39–103 · low confidence |
Data Center GPU Max 1550 Intel | 128 GB | 3,280 GB/s | Jan 2023 | 15.7 GB | Q8_0 | Comfortable |
|
63.1
tok/s
38–101 · low confidence |
Data Center GPU Max Subsystem Intel | 128 GB | 3,210 GB/s | Jan 2023 | 15.7 GB | Q8_0 | Comfortable |
|
61.7
tok/s
37–99 · low confidence |
A100 SXM4 80 GB NVIDIA | 80 GB | 2,040 GB/s | Nov 2020 | 15.7 GB | Q8_0 | Comfortable |
|
61.7
tok/s
37–99 · low confidence |
A100X NVIDIA | 80 GB | 2,040 GB/s | Jun 2021 | 15.7 GB | Q8_0 | Comfortable |
|
61.7
tok/s
37–99 · low confidence |
A800 SXM4 80 GB NVIDIA | 80 GB | 2,040 GB/s | Aug 2022 | 15.7 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
- NVIDIA
- Organisation type
- Industry
- Country
- United States of America
- Published
- 7 January 2025
- Authors
- NVIDIA: Niket Agarwal, Arslan Ali, Maciej Bala, Yogesh Balaji, Erik Barker, Tiffany Cai, Prithvijit Chattopadhyay, Yongxin Chen, Yin Cui, Yifan Ding, Daniel Dworakowski, Jiaojiao Fan, Michele Fenzi, Francesco Ferroni, Sanja Fidler, Dieter Fox, Songwei Ge, Yunhao Ge, Jinwei Gu, Siddharth Gururani, Ethan He, Jiahui Huang, Jacob Huffman, Pooya Jannaty, Jingyi Jin, Seung Wook Kim, Gergely Klár, Grace …
What it does
The problem areas the model was built for. A model can carry several of each.
- Domain
- Video, Vision, Robotics
- Task
- Robotic manipulation, System control, Video generation
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
- 14B
- Training data
- tokens
14B
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
- 2.8 × 10²⁴ FLOP
- How it was established
- Hardware
"We train all of the WFM models reported in the paper using a cluster of 10,000 NVIDIA H100 GPUs in a time span of three months." I assign the FLOPs from this cluster proportional to the parameter size of the model trained. There are a total of 76B parameters between the 8 models. Therefore, assuming 20% utilization (starting with 33% but then accounting for time between experiments), we get (10k H100s)*(90 days)*(24*60*60)*(979e12)*(0.2 utilization)*(14/76) = 2.8e24 FLOPs
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 H100 SXM5 80GB
- Chips used
- 10,000
- Power draw
- 13.8 MW
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
- Hugging Face
- nvidia
https://huggingface.co/nvidia/Cosmos-Predict1-14B-Video2World NVIDIA Open Model License Agreement Under the NVIDIA Open Model License, NVIDIA confirms: Models are commercially usable. You are free to create and distribute Derivative Models. NVIDIA does not claim ownership to any outputs generated using the Models or Derivative Models. Important Note: If you bypass, disable, reduce the efficacy of, or circumvent any technical limitation, safety guardrail or associated safety guardrail hyperpar…
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
- Cosmos World Foundation Model Platform for Physical AI
- Last updated
- 28 November 2025
The extremes
The ten fastest GPUs that run Cosmos-Predict1-14b-Video2World
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.
- 01 B300 288 GB · 8,000 GB/s · Q8_0 242 tok/s
- 02 B200 180 GB · 8,000 GB/s · Q8_0 242 tok/s
- 03 Radeon Instinct MI350X 288 GB · 8,190 GB/s · Q8_0 193 tok/s
- 04 Radeon Instinct MI355X 288 GB · 8,190 GB/s · Q8_0 193 tok/s
- 05 Radeon Instinct MI300 128 GB · 6,550 GB/s · Q8_0 155 tok/s
- 06 H200 NVL 141 GB · 4,890 GB/s · Q8_0 148 tok/s
- 07 H200 SXM 141 GB 141 GB · 4,890 GB/s · Q8_0 148 tok/s
- 08 Radeon Instinct MI325X 256 GB · 6,000 GB/s · Q8_0 142 tok/s
- 09 Radeon Instinct MI300A 128 GB · 5,325 GB/s · Q8_0 126 tok/s
- 10 Radeon Instinct MI300X 192 GB · 5,325 GB/s · Q8_0 126 tok/s
The smallest GPUs that still run Cosmos-Predict1-14b-Video2World
The cheapest route in, by memory capacity. A tight fit runs the model but leaves nothing spare for a longer conversation.
- 01 Arc B570 10 GB · needs 8.4 GB · IQ4_XS · tight 18.4 tok/s
- 02 Xbox Series X 6nm GPU 10 GB · needs 8.4 GB · IQ4_XS · tight 32.5 tok/s
- 03 Radeon RX 6750 GRE 10 GB 10 GB · needs 8.4 GB · IQ4_XS · tight 18.5 tok/s
- 04 CMP 170HX 10 GB 10 GB · needs 8.4 GB · IQ4_XS · tight 116 tok/s
- 05 CMP 90HX 10 GB · needs 8.4 GB · IQ4_XS · tight 56.5 tok/s
- 06 CMP 50HX 10 GB · needs 8.4 GB · IQ4_XS · tight 41.6 tok/s
- 07 Radeon RX 6700 10 GB · needs 8.4 GB · IQ4_XS · tight 18.5 tok/s
- 08 Radeon RX 6700M 10 GB · needs 8.4 GB · IQ4_XS · tight 18.5 tok/s
- 09 Xbox Series X GPU 10 GB · needs 8.4 GB · IQ4_XS · tight 32.5 tok/s
- 10 GeForce RTX 3080 10 GB · needs 8.4 GB · IQ4_XS · tight 56.5 tok/s
What the numbers mean
What it takes to run this model
Minimum card
P102-101
Memory needed
8.4 GB
Fastest
242 tok/s
Cosmos-Predict1-14b-Video2World is small enough at 14B parameters that hardware is rarely the obstacle — 306 of the cards we track can run it, including cards several years old.
The smallest card that holds it is the P102-101 with 10 GB, running it at IQ4_XS and producing around 20.2 tokens per second.
The quickest result comes from a B200 at around 242 tokens per second — its 8,000 GB/s of bandwidth is what buys that.
What this model is
Cosmos-Predict1-14b-Video2World was published by NVIDIA, in United States of America, in January 2025. It comes out of industry.
It works in Video, Vision, Robotics, and is recorded as doing robotic manipulation, System control, Video generation.
Published weights mean the model runs on your machine rather than someone else's, which is what makes the hardware question below answerable at all. It is published under the nvidia organisation on Hugging Face.
What decides the speed
Half the cards that hold it manage more than 20.4 tokens per second, and 268 exceed reading speed outright.
It is a dense model, so every parameter is read for every token produced. That makes speed track memory bandwidth almost exactly — a card with twice the bandwidth generates roughly twice as fast.
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.
Training and provenance
The training run consumed about 2.8 × 10²⁴ FLOP, on NVIDIA H100 SXM5 80GB. That figure describes the cost of creating it and has no bearing on how quickly it generates text.
Step by step
How to choose a GPU for Cosmos-Predict1-14b-Video2World
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
The table lists every card that can hold Cosmos-Predict1-14b-Video2World — around 8.4 GB at IQ4_XS. That figure, not the card's headline performance, is what decides whether it runs.
-
02
Decide how long your conversations run
Longer conversations cost memory on top of what the weights need. Move the slider to your real working length before trusting any row for Cosmos-Predict1-14b-Video2World.
-
03
Choose how far you will compress it
Compression is what makes Cosmos-Predict1-14b-Video2World fit smaller cards, at some cost in accuracy — IQ4_XS on the smallest card that fits. A minimum quality removes the ones that go too far.
-
04
Sort by speed
The speed ordering for Cosmos-Predict1-14b-Video2World is effectively an ordering by memory bandwidth, which is why the B200 tops it at 242 tok/s.
-
05
Read the fit column last
The fit column separates cards that just manage Cosmos-Predict1-14b-Video2World from those with room to spare. Buy for the second if the context might grow.
-
06
Check the card from the other side
Following a card through to its own page shows every other model it can hold, which is the question that follows once Cosmos-Predict1-14b-Video2World is settled.
Answers
Cosmos-Predict1-14b-Video2World — common questions
Would two GPUs run Cosmos-Predict1-14b-Video2World faster?
Capacity adds across cards; throughput does not. Since 306 of the cards we track already hold Cosmos-Predict1-14b-Video2World on their own, a second card is rarely the answer here.
Why does the quantisation differ between cards for Cosmos-Predict1-14b-Video2World?
A larger card holds a more accurate copy. Across the cards that run Cosmos-Predict1-14b-Video2World, 5 compression levels are used; the floor control above pins it to one.
How accurate are these Cosmos-Predict1-14b-Video2World speed estimates?
These are estimates with real error bars. The fastest result here, 145–387 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 Cosmos-Predict1-14b-Video2World?
The smallest card in our catalogue that holds Cosmos-Predict1-14b-Video2World is the P102-101, with 10 GB of memory. It runs the model at IQ4_XS using about 8.4 GB, and produces roughly 20.2 tokens per second. 306 cards in total can run it.
How fast is Cosmos-Predict1-14b-Video2World on a GPU?
It depends on the card. The quickest we calculate is a B200 at about 242 tokens per second; the slowest that still runs it manages considerably less. Reading speed is around ten tokens per second, and 268 of the cards that can run Cosmos-Predict1-14b-Video2World clear that.
How much VRAM does Cosmos-Predict1-14b-Video2World need?
About 8.4 GB at IQ4_XS 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 Cosmos-Predict1-14b-Video2World on a 12 GB GPU?
Yes. A GeForce RTX 3080 Ti with 12 GB runs it at Q5_K_M, using about 10.8 GB and generating roughly 49.3 tokens per second — a tight fit.
Can I run Cosmos-Predict1-14b-Video2World on a 16 GB GPU?
Yes. A Tesla V100 SXM2 16 GB with 16 GB runs it at Q6_K, using about 12.4 GB and generating roughly 49.7 tokens per second — a tight fit.
Can I run Cosmos-Predict1-14b-Video2World on a 24 GB GPU?
Yes. A GeForce RTX 5090 D V2 with 24 GB runs it at Q8_0, using about 15.7 GB and generating roughly 40.5 tokens per second — a comfortable fit.
Is Cosmos-Predict1-14b-Video2World open source?
Its weights are published, so Cosmos-Predict1-14b-Video2World 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 Cosmos-Predict1-14b-Video2World have?
Cosmos-Predict1-14b-Video2World has 14B parameters. 14B. 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 Cosmos-Predict1-14b-Video2World?
Cosmos-Predict1-14b-Video2World was published by NVIDIA, based in United States of America, categorised as industry.
When was Cosmos-Predict1-14b-Video2World released?
Cosmos-Predict1-14b-Video2World was published in January 2025.
What is Cosmos-Predict1-14b-Video2World used for?
Cosmos-Predict1-14b-Video2World works in Video, Vision, Robotics, and is recorded as handling robotic manipulation, System control, Video generation. A model can carry several of each, so these are the areas it was built for rather than a limit on what it will attempt.
Where can I download Cosmos-Predict1-14b-Video2World?
Its weights are published under the nvidia 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 Cosmos-Predict1-14b-Video2World?
Around 2.8 × 10²⁴ FLOP, on NVIDIA H100 SXM5 80GB. 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 Cosmos-Predict1-14b-Video2World if it does not fit in my GPU?
Only by offloading, which is usually a false economy: the part in system memory drags the whole thing down — the nearest miss we calculate is short by 2.0 GB. Our figures for Cosmos-Predict1-14b-Video2World assume it is fully resident.
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.