Best local AI models for NVIDIA MX130

The largest models that fit fully

The 30 largest of the 56 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 MX130 is an entry level laptop graphics card equipped with 2 GB of GDDR5 dedicated video memory. This hardware memory limit dictates the size of the artificial intelligence models you can run locally on your system. To execute a model entirely on your graphics processor, the active weights must fit within this 2 GB boundary. Running models directly on your video memory ensures the fastest possible processing speeds for your tasks.

The quantization column indicates the compression level applied to each model. Quantization reduces the precision of model weights to save space. For example, the Allegro 2.8B model uses a Q4_K_M quantization to fit into 2 GB of video memory. Models like SmolVLM 2B use a higher quality Q6_K quantization which fits exactly into 2 GB. Smaller models like TinyLlama 1.1B can run at the highest Q8_0 quantization while using only 1.4 GB of video memory.

When a model exceeds your 2 GB video memory, you must use CPU offload. This technique splits the model weights between your graphics card and your system RAM. We assume your laptop has 32 GB of system RAM for these scenarios. For example, running the SmolLM3 3B model at Q4_K_M requires 2.2 GB of video memory and 4.2 GB of system RAM. Offloading allows you to run larger models like MusicGen at 3.3B or SDXL Turbo at 3.5B, but your processing speed will decrease significantly.

Be aware of the context window limits when running local language models on tight hardware. The memory usage figures listed are calculated at a standard 4k context window. If you increase the context length to process longer documents or chat histories, the memory consumption of the key value cache will grow. This extra memory demand can quickly push a model past your 2 GB limit and trigger slow CPU offloading.

You can choose from several optimized models that fit your hardware constraints. For image generation, Stable Diffusion 3.5 Medium fits at Q4_K_M quantization while using 1.8 GB of video memory. For audio and speech tasks, Whisper Large v3 runs at Q8_0 quantization using 2 GB of video memory. For text generation, Qwen3 1.7B fits comfortably at Q6_K quantization using 1.7 GB of video memory.