Best local AI models for AMD R5 M315

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 AMD Radeon R5 M315 is an entry level graphics card equipped with 2 GB of DDR3 video memory. This limited VRAM buffer dictates which artificial intelligence models can run directly on your hardware. To execute a model entirely on the graphics processor without system memory bottlenecks, the total size of the model must remain under this 2 GB threshold.

Model quantization is a compression technique that reduces the precision of model weights to save space. The best quant column indicates the highest quality quantization level that still fits within the hardware limits. For example, the 2.8B Allegro model and the 2.7B Open-Sora Plan model can run on this card using the Q4_K_M quantization format, which uses exactly 2 GB of video memory.

As model sizes decrease, you can utilize higher quality quantization formats. The 2.3B SeamlessM4T v2 and the 2.2B Parler-TTS models run at the Q5_K_M quantization level. Smaller models like the 2B SmolVLM, 2B Stable Diffusion 3 Medium, and 1.9B Moondream 2 can utilize the Q6_K format. The smallest models, including the 1.6B StableLM 2, 1.55B Whisper Large v3, and 1.1B TinyLlama, can run at the premium Q8_0 quantization level.

When a model exceeds the 2 GB VRAM limit, you must use CPU offloading. This process splits the workload between your graphics card and your system RAM. For instance, running the 3B SmolLM3 or the 3B Kandinsky 3.1 requires 2.2 GB of video memory at Q4_K_M, which forces 4.2 GB of data into your system RAM. Larger models like the 3.417B Stable Diffusion XL require 4.1 GB of video memory at FP8 and 6.1 GB of system RAM.

CPU offloading allows you to run larger architectures, but it introduces a severe performance penalty. System DDR3 or DDR4 RAM is significantly slower than dedicated video memory. This transfer bottleneck will result in much slower generation speeds. Additionally, running text models with a standard 4k context window increases memory usage during active inference, which can cause out of memory errors if your VRAM is already fully allocated.