Best local AI models for NVIDIA GT 730M

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 GT 730M is an entry level mobile graphics card equipped with 2 GB of DDR3 video memory. This hardware memory limit dictates the maximum size of the artificial intelligence models you can run locally. To load a model entirely on this GPU, the total memory footprint of the model must remain under the 2 GB threshold. This page lists the compatible models that fit within these tight hardware constraints.

To make larger models fit into the 2 GB video memory, developers use quantization. The quant column indicates the specific level of compression applied to the model weights. For example, the 2.8B Allegro model fits using the Q4_K_M quant which uses exactly 2 GB of video memory. Smaller models like the 1.7B SmolLM2 or 1.7B Qwen3 can run at a higher quality Q6_K quant while using 1.7 GB of video memory. Models like the 1.1B TinyLlama can run at the high quality Q8_0 quant using 1.4 GB of video memory.

When a model exceeds the 2 GB video memory of the GT 730M, you must use CPU offload. This technique splits the model layers between your graphics card and your system RAM. We assume your computer has 32 GB of system RAM for these calculations. For instance, running the 3B SmolLM3 or 3B Replit Code v1.5 requires 2.2 GB of video memory at Q4_K_M along with 4.2 GB of system RAM. CPU offload allows you to run larger models but it significantly reduces processing speed.

Using CPU offload also enables image generation models like Stable Diffusion XL. This 3.417B model needs 4.1 GB of memory at FP8 or optimized settings which requires 6.1 GB of system RAM to assist the GPU. Similarly, the 3.5B SDXL Turbo and 3.5B SDXL Lightning models need 2.6 GB at Q4_K_M which requires 4.6 GB of system RAM. While these models can run through offloading, the generation times will be slow due to the DDR3 memory bandwidth.

You must also consider the memory cost of context length. The memory figures listed here represent the base model size. Running text models with a standard 4k context window requires additional video memory to store the active conversation history. On a 2 GB card, this extra memory requirement can easily cause an out of memory error. You may need to reduce your context window size to keep the model running entirely on the GPU.