Best local LLMs for 12 GB VRAM
Every catalogued model that fits entirely in 12 GB, with no layers moved to system RAM. Speeds are estimated on the NVIDIA GeForce RTX 3060 12GB, the most common 12 GB card in Steam's hardware survey.
- Largest popular model that fitsQwen2.5-Coder-14B-Instruct15B params · Q4_K_M · needs 11.4 GB~24 tok/s About reading paceOpen in the calculator →
- Best fast pickLLaDA2.0-mini16B params · Q4_K_M · needs 11.0 GB~229 tok/s Faster than you readOpen in the calculator →
- Most downloaded that fitsQwen3.5-9B9.7B params · Q4_K_M · needs 7.05 GB~50 tok/s Faster than you readOpen in the calculator →
“Best” is not a quality ranking. The picks are the largest widely downloaded model that fits, the largest that still answers at 30 tokens a second or more, and the most downloaded; the list puts current, widely downloaded releases first. Speeds are estimates from memory bandwidth, not benchmarks.
Is 12 GB of VRAM enough for a local LLM?
For models under 16B parameters, yes: all 145 in the catalogue fit entirely at Q4_K_M, the format most people download. In the 16B to 36B band, 1 of 66 fits, and nothing larger does. 136 of the 147 that fit answer at 30 tokens a second or more on the NVIDIA GeForce RTX 3060 12GB, which is faster than most people read. A longer conversation needs more memory than the 8,192 tokens counted here, and a smaller format than Q4_K_M needs less; the calculator sizes either.
| Model size | Fit in 12 GB | For example |
|---|---|---|
| Under 4B parameters | 66 of 66 | Qwen3.5-2B · needs 2.32 GB |
| 4B to 9B parameters | 58 of 58 | Qwen3.5-4B · needs 3.98 GB |
| 9B to 16B parameters | 21 of 21 | Qwen3.5-9B · needs 7.05 GB |
| 16B to 36B parameters | 1 of 66 | LLaDA2.0-mini · needs 11.0 GB |
| 36B and larger parameters | 0 of 51 | none fits entirely |
Models that fit in 12 GB
| Model | Parameters | Needs | Spare | Decode | Answer |
|---|---|---|---|---|---|
| Qwen3.5-9B Qwen · Q4_K_M | 9.7B | 7.05 GB | 4.95 GB | ~50 tok/s | Details |
| Qwen3.5-4B Qwen · Q4_K_M | 4.7B | 3.98 GB | 8.02 GB | ~91 tok/s | Details |
| gemma-4-12B-it Google · Q4_K_M | 12B | 9.54 GB | 2.46 GB | ~30 tok/s | Details |
| Qwen3.5-2B Qwen · Q4_K_M | 2.3B | 2.32 GB | 9.68 GB | ~198 tok/s | Details |
| gemma-4-E4B-it Google · Q4_K_M | 8.0B | 5.86 GB | 6.14 GB | ~81 tok/s | Details |
| NVIDIA-Nemotron-3-Nano-4B-BF16 NVIDIA · Q4_K_M | 4.0B | 3.46 GB | 8.54 GB | ~92 tok/s | Details |
| gemma-4-E2B-it Google · Q4_K_M | 5.1B | 4.02 GB | 7.98 GB | ~214 tok/s | Details |
| Qwen3.5-0.8B Qwen · Q4_K_M | 873M | 1.46 GB | 10.5 GB | ~435 tok/s | Details |
| Qwen3-VL-8B-Instruct Qwen · Q4_K_M | 8.8B | 7.04 GB | 4.96 GB | ~42 tok/s | Details |
| North-Micro-Vision-Instruct CohereLabs · Q4_K_M | 2.5B | 2.99 GB | 9.01 GB | ~131 tok/s | Details |
| granite-4.2-8b IBM · Q4_K_M | 8.8B | 7.49 GB | 4.51 GB | ~38 tok/s | Details |
| granite-4.1-3b IBM · Q4_K_M | 3.4B | 3.72 GB | 8.28 GB | ~88 tok/s | Details |
| LFM2.5-2.6B LiquidAI · Q4_K_M | 2.7B | 2.61 GB | 9.39 GB | ~142 tok/s | Details |
| Qwen3-VL-4B-Instruct Qwen · Q4_K_M | 4.4B | 4.72 GB | 7.28 GB | ~66 tok/s | Details |
| Qwen3-VL-2B-Instruct Qwen · Q4_K_M | 2.1B | 3.05 GB | 8.95 GB | ~120 tok/s | Details |
| granite-4.2-3b IBM · Q4_K_M | 3.7B | 3.72 GB | 8.28 GB | ~88 tok/s | Details |
| LFM2.5-230M LiquidAI · Q4_K_M | 230M | 1.05 GB | 10.9 GB | ~988 tok/s | Details |
| Qwen3-0.6B Qwen · Q4_K_M | 752M | 2.22 GB | 9.78 GB | ~184 tok/s | Details |
| granite-4.1-8b IBM · Q4_K_M | 8.8B | 7.49 GB | 4.51 GB | ~38 tok/s | Details |
| LFM2.5-VL-3B LiquidAI · Q4_K_M | 3.1B | 2.85 GB | 9.15 GB | ~142 tok/s | Details |
| MiMo-V2.6-Distill-Qwen-9B XiaomiMiMo · Q4_K_M | 9.4B | 6.90 GB | 5.10 GB | ~50 tok/s | Details |
| Qwen3-4B-Instruct-2507 Qwen · Q4_K_M | 4.0B | 4.72 GB | 7.28 GB | ~66 tok/s | Details |
| Ling-3.0-tiny inclusionAI · Q4_K_M | 7.9B | 5.85 GB | 6.15 GB | ~285 tok/s | Details |
| Qwen3-8B Qwen · Q4_K_M | 8.2B | 7.04 GB | 4.96 GB | ~42 tok/s | Details |
| Olmo-3-7B-Instruct allenai · Q4_K_M | 7.3B | 7.96 GB | 4.04 GB | ~35 tok/s | Details |
| Qwen3-4B Qwen · Q4_K_M | 4.0B | 4.51 GB | 7.49 GB | ~66 tok/s | Details |
| LFM2.5-8B-A1B LiquidAI · Q4_K_M | 8.5B | 6.06 GB | 5.94 GB | ~221 tok/s | Details |
| LFM2.5-350M LiquidAI · Q4_K_M | 354M | 1.13 GB | 10.9 GB | ~638 tok/s | Details |
| Hy-MT2-1.8B Tencent · Q4_K_M | 2.0B | 2.47 GB | 9.53 GB | ~147 tok/s | Details |
| LLaDA2.0-mini inclusionAI · Q4_K_M | 16B | 11.0 GB | 1.00 GB | ~229 tok/s | Details |
| LFM2.5-1.2B-Instruct LiquidAI · Q4_K_M | 1.2B | 1.63 GB | 10.4 GB | ~235 tok/s | Details |
| Ministral-3-14B-Instruct-2512 Mistral AI · Q4_K_M | 14B | 10.4 GB | 1.62 GB | ~27 tok/s | Details |
| Qwen3-1.7B Qwen · Q4_K_M | 2.0B | 3.02 GB | 8.98 GB | ~120 tok/s | Details |
| Nemotron-3.5-Content-Safety NVIDIA · Q4_K_M | 4.3B | 3.73 GB | 8.27 GB | ~88 tok/s | Details |
| Hy-MT2-7B Tencent · Q4_K_M | 8.0B | 6.50 GB | 5.50 GB | ~43 tok/s | Details |
| Qwen3-14B Qwen · Q4_K_M | 15B | 11.1 GB | 0.86 GB | ~25 tok/s | Details |
| GLM-4.6V-Flash zai-org · Q4_K_M | 10B | 7.45 GB | 4.55 GB | ~43 tok/s | Details |
| Qwen2.5-VL-7B-Instruct Qwen · Q4_K_M | 8.3B | 6.36 GB | 5.64 GB | ~51 tok/s | Details |
| SmolLM3-3B HuggingFaceTB · Q4_K_M | 3.1B | 3.46 GB | 8.54 GB | ~97 tok/s | Details |
| Ministral-3-8B-Instruct-2512 Mistral AI · Q4_K_M | 8.9B | 7.14 GB | 4.86 GB | ~41 tok/s | Details |
14 devices with 12 GB
The same models fit on every one of them. What changes is speed, which follows memory bandwidth: a card with twice the bandwidth decodes roughly twice as fast.
Which card to buy, at every memory size →
Store links may pay us a commission. They never decide which models are listed — the memory arithmetic does.
What does not fit in 12 GB
Widely downloaded models that need more than 12 GB at Q4_K_M, and the smallest memory size that holds each one entirely. Each link shows what 12 GB can still do with it: a smaller format, or part of the model in system RAM at a lower speed.
| Model | Needs | Fits from | On 12 GB |
|---|---|---|---|
| gpt-oss-20b | 13.8 GB | 16 GB | On 12 GB |
| gpt-oss-safeguard-20b | 13.8 GB | 16 GB | On 12 GB |
| Qwen3.8-27B | 18.6 GB | 20 GB | On 12 GB |
| gemma-4-26B-A4B-it | 16.8 GB | 20 GB | On 12 GB |
| Qwen3.6-27B | 18.6 GB | 20 GB | On 12 GB |
| Qwen3.5-27B | 18.6 GB | 20 GB | On 12 GB |
6 more models fit in 16 GB. Best local LLMs for 16 GB →