Best local AI models for NVIDIA RTX 5090 Laptop

The largest models that fit fully

The 30 largest of the 173 models that fit; every smaller model in the catalog fits too. Best quant means the highest quality compression whose weights and 4k context both sit inside the memory.

Close, but only with CPU offload

These need more than the card holds at their smallest practical quant, so part of the model runs from system memory (figures assume 32 GB of it). They work, several times slower.

How to read this

The NVIDIA RTX 5090 Laptop GPU features 24 GB of GDDR7 memory. This dedicated memory determines the size of the artificial intelligence models you can run locally. To run a model entirely on your graphics hardware, the total size of the model must remain under this 24 GB limit. This page lists the largest compatible models and their optimal configurations.

The quantization column shows the compression level applied to each model. Quantization reduces the size of the model weights to save memory. For example, Qwen3 32B, Qwen3.5 32B, Aya Expanse 32B, Granite 4.0 32B, and Qwen2.5-Coder 32B fit within 23.4 GB using the Q4_K_M quantization. Similarly, Qwen3-30B-A3B and Qwen3-Coder 30B-A3B require 22 GB at Q4_K_M.

Higher quantization levels preserve more model intelligence but require more memory. Gemma 3 27B, Gemma 3 27B vision, and Wan 2.2 / 2.5 use 23 GB at the Q5_K_M level. Gemma 4 26B-A4B and Gemma 4 all sizes fit within 22.2 GB at Q5_K_M. Aria uses 21.3 GB at Q5_K_M. Mistral Small 3.2, Magistral Small, and Devstral Small 1.1 run at Q6_K using 23.6 GB. Solar Pro and Codestral 22B use 21.6 GB at Q6_K. The gpt-oss-20b and Reka Flash 3 models use 20.7 GB at Q6_K. Qwen-Image and Qwen-Image-Edit use 19.7 GB at Q6_K. CogVLM2 uses 18.7 GB at Q6_K.

For maximum precision, you can run smaller models at the Q8_0 quantization level. HunyuanImage 2.1 / 3.0 uses 21.6 GB. Ling-Coder-Lite uses 21.4 GB. DeepSeek-Coder-V2 16B and Kimi-VL A3B use 20.4 GB. Apriel-1.5-15B-Thinker and StarCoder2 15B use 19.1 GB. Qwen2.5 14B uses 19.5 GB. All these configurations run entirely on your graphics hardware.

When a model exceeds the 24 GB limit, you must offload parts of it to your system RAM. This offloading process reduces generation speed because system memory is slower than GDDR7. For example, OTel 2.0 LLM 31B IT needs 27.5 GB at Q4_K_M and requires 29.5 GB of system RAM. DeepSeek-Coder 33B and WizardCoder 33B need 24.2 GB at Q4_K_M and require 26.2 GB of system RAM. Yi 1.5 34B, Granite Code 34B, LLaVA 1.5 / 1.6 34B, and Ovis 2 need 24.9 GB at Q4_K_M and require 26.9 GB of system RAM. Qwen3.6-35B-A3B and Command R 35B need 25.6 GB at Q4_K_M and require 27.6 GB of system RAM. Seed-OSS 36B needs 26.4 GB at Q4_K_M and requires 28.4 GB of system RAM.

All memory calculations on this page assume a standard 4k context window. Running longer conversations or processing larger documents increases memory consumption. If you increase the context length beyond 4000 tokens, the model will require more memory and you may need to choose a smaller model size or a lower quantization level.