Best local AI models for AMD R5 M335

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 M335 is an entry level laptop graphics card equipped with 2 GB of DDR3 video memory. This 2 GB VRAM limit defines the maximum size of any AI model you can run entirely on the GPU. Because DDR3 memory has lower bandwidth than modern GDDR graphics memory, keeping the entire model inside VRAM is critical for maintaining usable processing speeds.

To fit models into this 2 GB limit, you must use quantized versions. Quantization reduces the precision of model weights to save space. The best quant column shows the optimal balance of size and quality for this hardware. For example, the 2.8B Allegro model and the 2.7B Open-Sora Plan model both fit into 2 GB of VRAM when using the Q4_K_M quantization level. Similarly, the 2.6B Playground v2.5 and the 2.5B Stable Diffusion 3.5 Medium fit using Q4_K_M while consuming 1.9 GB and 1.8 GB of VRAM respectively.

Smaller models can use higher precision quants for better output quality. The 2B Stable Diffusion 3 Medium and the 2B SmolVLM 256M / 500M / 2B both run at the Q6_K level while utilizing the full 2 GB of VRAM. Extremely compact models like the 1.6B StableLM 2 1.6B and the 1.1B TinyLlama 1.1B can run at the Q8_0 level. These Q8_0 quants use 2 GB and 1.4 GB of VRAM respectively, offering maximum fidelity within the hardware limits.

When a model exceeds the 2 GB VRAM limit, you must use CPU offloading. This process splits the model layers between your GPU and your system RAM. For these scenarios, we assume your computer has 32 GB of system RAM. Offloading allows you to run larger models like the 3B SmolLM3 3B or the 3.417B Stable Diffusion XL. However, offloading comes with a massive performance cost because data must travel over the slow system bus instead of staying on the graphics card.

For instance, running the 3B SmolLM3 3B at Q4_K_M requires 2.2 GB of VRAM and 4.2 GB of system RAM. Running Stable Diffusion XL at FP8 requires 4.1 GB of VRAM and 6.1 GB of system RAM. While offloading makes these larger models run, the processing speed will be significantly slower than models that fit entirely within the 2 GB VRAM limit.

You must also consider the 4k context window caveat when running text models. The memory figures listed only cover the base model weights. As you type longer prompts and the model generates longer responses, the active context memory grows. Running a model near the 2 GB limit with a full 4k context window will likely exceed your VRAM and force slow system RAM usage.