OpenModelsHQ

GeForce RTX 5090

NVIDIA · Consumer GPU

GeForce RTX 5090 has 32 GB of memory and 1,792 GB/s of memory bandwidth. 391 open models in our directory fit at 4-bit.

Usable for models, before runtime overhead
32GB
Listed memory, before overhead
Total memory
32 GB
Bandwidth
1,792 GB/s
Models that fit at 4-bit
391
Within 28.8 GB, 90% of usable

Planning a multi-GPU deployment? Check NVLink domains and network topology as well as memory fit. A GPU specification alone does not establish the interconnect in a server or cloud allocation.

What fits by category
On GeForce RTX 5090 at 4-bit, 391 models fit, 0 are tight and 100 are too big, across 6 categories. Fits means the estimate is within 90% of usable memory; tight is between 90% and 100%. Models without a published size are left out.
What fits by category: data
CategoryFitsTightToo bigTotal
LLM208072280
Vision79027106
Image230124
Video220022
Speech400040
Embedding190019

Largest open models that fit at 4-bit

See all 391 in the table

A model fits when its 4-bit memory estimate (8k context) stays within 90% of usable memory, about 28.8 GB here. Showing the 30 largest of 391.

ModelTypeSizeMemory at 4-bit
LLM40.4B28.5 GB
LLM40.4B28.5 GB
LLM40.4B28.5 GB
LLM40.4B28.5 GB
LLM40.4B28.5 GB
LLM36.2B26.2 GB
LLM36.2B26.2 GB
LLM35B24.9 GB
Vision33.3B24.6 GB
LLM31.2B24.1 GB
LLM32.8B24 GB
Speech33.7B24 GB
LLM32.8B24 GB
LLM32.8B24 GB
LLM32.8B24 GB
LLM32.8B24 GB
Vision36B23.9 GB
LLM32.9B23.5 GB
Vision34.4B23.3 GB
LLM32.3B23 GB
Image32.2B23 GB
LLM32B22.8 GB
LLM32B22.8 GB
LLM27.4B22.7 GB
LLM30B22.2 GB
LLM32.9B22.2 GB
Vision31.3B22.1 GB
LLM32.6B22.1 GB
LLM32B21.7 GB
LLM32B21.7 GB

These are memory checks, not speed or runtime compatibility results.

Compared with

All hardware
HardwareTotal memoryUsable for modelsBandwidth
GeForce RTX 5090This device
32 GBUp to 32 GB1,792 GB/s
24 GBUp to 24 GB960 GB/s
GeForce RTX 3090Discontinued
24 GBUp to 24 GB936 GB/s
GeForce RTX 4090Discontinued
24 GBUp to 24 GB1,008 GB/s

Capacity and bandwidth answer different questions: memory decides what fits, bandwidth drives how fast it runs.

Launch dates, useful-life evidence and ownership cost scenarios