Dedicated GPU memory guide
Best local AI models for 8GB VRAM
This guide uses a Windows NVIDIA profile with 8GB dedicated GPU VRAM. It separates full GPU fits, tight fits, and partial offload instead of treating model file size as an official memory requirement.
Dedicated VRAM is not the same as system RAM or Apple Silicon unified memory. For Mac planning, use the Apple Silicon guides so the fit engine uses unified-memory terminology and a Mac-specific conservative budget.
Likely full GPU fit
Good fit
Llama 3.2 3B
Llama 3.2 3B · Q4_K_M · 3.7 GB planning memory · 2.7 GB minimum load
Good fit
Phi-3.5 Mini
Phi-3.5 Mini · Q4_K_M · 4.2 GB planning memory · 3.2 GB minimum load
Good fit
Phi-4 Mini
Phi-4 Mini · Q4_K_M · 4.2 GB planning memory · 3.2 GB minimum load
Good fit
Gemma 3 4B
Gemma 3 4B · Q4_K_M · 4.5 GB planning memory · 3.5 GB minimum load
Good fit
Qwen3 4B
Qwen3 4B · Q4_K_M · 4.5 GB planning memory · 3.5 GB minimum load
Good fit
Llama 3.2 3B
Llama 3.2 3B · Q8_0 · 5.2 GB planning memory · 4.2 GB minimum load
Good fit
DeepSeek-R1 Distill Qwen 7B
DeepSeek-R1 Distill Qwen 7B · Q4_K_M · 6.1 GB planning memory · 5.1 GB minimum load
Good fit
Mistral 7B
Mistral 7B · Q4_K_M · 6.1 GB planning memory · 5.1 GB minimum load
Good fit
Gemma 3 4B
Gemma 3 4B · Q8_0 · 6.2 GB planning memory · 5.2 GB minimum load
Good fit
Qwen3 4B
Qwen3 4B · Q8_0 · 6.2 GB planning memory · 5.2 GB minimum load
Good fit
Llama 3.1 8B
Llama 3.1 8B · Q4_K_M · 6.6 GB planning memory · 5.6 GB minimum load
Good fit
Qwen3 8B
Qwen3 8B · Q4_K_M · 7 GB planning memory · 6 GB minimum load
Good fit
Llama 3.1 8B
Llama 3.1 8B · Q5_K_M · 7.8 GB planning memory · 6.8 GB minimum load
Good fit
Qwen3 8B
Qwen3 8B · Q5_K_M · 7.8 GB planning memory · 6.8 GB minimum load
Tight fit
Partial offload
Partial offload
DeepSeek-R1 Distill Qwen 14B
DeepSeek-R1 Distill Qwen 14B · Q4_K_M · 10.9 GB planning memory · 9.3 GB minimum load
Partial offload
Phi-4
Phi-4 · Q4_K_M · 10.9 GB planning memory · 9.3 GB minimum load
Partial offload
Gemma 3 12B
Gemma 3 12B · Q5_K_M · 11.6 GB planning memory · 9.9 GB minimum load
Partial offload
Qwen3 14B
Qwen3 14B · Q4_K_M · 12.2 GB planning memory · 10.4 GB minimum load
Partial offload
Qwen3 14B
Qwen3 14B · Q5_K_M · 13.6 GB planning memory · 11.6 GB minimum load
Partial offload
Mistral Small 3.2 24B
Mistral Small 3.2 24B · Q4_K_M · 20.3 GB planning memory · 17.3 GB minimum load
Partial offload
Gemma 3 27B
Gemma 3 27B · Q4_K_M · 22.3 GB planning memory · 19 GB minimum load
Partial offload
Qwen3.8 27B
Qwen3.8 27B · Q4_K_M · 24.3 GB planning memory · 20.7 GB minimum load
Partial offload
Qwen3 30B-A3B
Qwen3 30B-A3B · Q4_K_M · 25.7 GB planning memory · 21.9 GB minimum load
Partial offload
Gemma 3 27B
Gemma 3 27B · Q5_K_M · 26.2 GB planning memory · 22.3 GB minimum load
Partial offload
DeepSeek-R1 Distill Qwen 32B
DeepSeek-R1 Distill Qwen 32B · Q4_K_M · 27 GB planning memory · 23 GB minimum load