Best LLMs for 6GB VRAM
With 6 GB of VRAM you can run models up to Gemma 4 E4B (8B) at Q4_K_M with an 8K context. 13 of 58 tracked models fit fully, and 5 more fit at a lower quantization.
Models that fit in 6 GB
| Model | VRAM at 8K | Best quant that fits | Max context at Q4 |
|---|---|---|---|
| Gemma 4 E4B | 5.7 GB | IQ4_XS | 4K |
| Qwen 2.5 7B | 5.6 GB | Q4_K_M | 9K |
| DeepSeek R1 Distill Qwen 7B | 5.6 GB | Q4_K_M | 9K |
| Mistral 7B v0.3 | 5.7 GB | IQ4_XS | 7K |
| Gemma 4 E2B | 4.0 GB | Q6_K | 128K |
| Gemma 3 4B | 3.7 GB | Q8_0 | 105K |
| Qwen 3.5 4B | 3.7 GB | Q8_0 | 70K |
| Qwen 3 4B | 4.2 GB | Q6_K | 18K |
| Phi-4 Mini | 4.1 GB | Q6_K | 20K |
| Ministral 3 3B | 3.5 GB | Q8_0 | 28K |
| Llama 3.2 3B | 3.5 GB | Q8_0 | 27K |
| Llama 3.2 1B | 1.7 GB | FP16 / BF16 | 128K |
| Gemma 3 1B | 1.6 GB | FP16 / BF16 | 32K |
Fits at a lower quantization
These need more than 6 GB at Q4_K_M but fit at a smaller format, with more quality loss.
| Model | VRAM at Q4_K_M | Fits at |
|---|---|---|
| Qwen 3.5 9B | 6.4 GB | Q3_K_M |
| Ministral 3 8B | 6.7 GB | Q2_K |
| Qwen 3 8B | 6.6 GB | Q2_K |
| Llama 3.1 8B | 6.3 GB | Q3_K_M |
| DeepSeek R1 Distill Llama 8B | 6.3 GB | Q3_K_M |
Image and video models
Best weight precision that fits 6 GB with model swapping.
| Model | Best precision |
|---|---|
| Stable Diffusion 1.5 | FP16 / BF16 |
| SDXL 1.0 | FP8 |
| Stable Diffusion 3.5 Medium | Needs offloading or more VRAM |
| Stable Diffusion 3.5 Large | Needs offloading or more VRAM |
| FLUX.1 [dev] | Needs offloading or more VRAM |
| FLUX.1 [schnell] | Needs offloading or more VRAM |
| Qwen-Image | Needs offloading or more VRAM |
| Wan 2.2 A14B (T2V / I2V) | Needs offloading or more VRAM |
| Wan 2.2 TI2V 5B | Needs offloading or more VRAM |
| HunyuanVideo | Needs offloading or more VRAM |
| FLUX.1 Kontext [dev] | Needs offloading or more VRAM |
| FLUX.2 [dev] | Needs offloading or more VRAM |
| Z-Image Turbo | GGUF Q5_K_S |
| Qwen-Image-Edit 2509 | Needs offloading or more VRAM |
| LTX-2 | Needs offloading or more VRAM |
6 GB VRAM: common questions
Is 6 GB of VRAM enough for local LLMs?
Yes for models up to about 8B at Q4_K_M with 8K context. 6 GB fits 13 of the 58 models tracked here, including Gemma 4 E4B, Qwen 2.5 7B, DeepSeek R1 Distill Qwen 7B.
What is the biggest model 6 GB can run?
Gemma 4 E4B (8B) at Q4_K_M with 8K context, using about 5.7 GB. Qwen 3.5 9B also fits at Q3_K_M, with some quality loss.
Is 6 GB enough for Flux and SDXL?
SDXL: FP8. FLUX.1 dev: Needs offloading or more VRAM (best precision that fits with model swapping).