RTX 3060 Laptop GPU — what LLMs can it run?
46 of 151 indexed models fit comfortably in 6GB at 4K context, each at the highest-quality quant that still leaves headroom.
The short answer for RTX 3060 Laptop GPU
6 GB, GDDR6, bandwidth not listed. 46 of 151 models in this index fit comfortably at 4K context; 54 load at all.
Hardware specifications source- Biggest that fits
- GLM-4-9B-Chat — 4.9 GB at AWQ INT4 — 1.1 GB spare at 4K, so longer context comes out of a thin margin.
- Room to grow
- Qwen2.5 3B Instruct — 3.6 GB at Q8_0 — under 60% of the card, which leaves 2.4 GB for a long context window or a second process.
- The wall you will hit
- Laptop cooling and power limits affect performance even between laptops with the same GPU. Use the listed VRAM for fit, not desktop benchmark results for speed.
4–10B · 24
| Model | Quant | Est. VRAM | Headroom |
|---|---|---|---|
| GLM-4-9B-Chat9B | AWQ INT4 | 4.87 GB | +1.1 GB |
| MiMo-V2.6-Distill-Qwen-9B9B | Q3_K_M | 4.82 GB | +1.2 GB |
| Clairic Flash 9B9B | Q3_K_M | 5.04 GB | +1 GB |
| Qwen3-VL 8B Instruct8B | AWQ INT4 | 5.22 GB | +0.8 GB |
| Qwen2-VL 7B Instruct7B | AWQ INT4 | 4.57 GB | +1.4 GB |
| Granite 3.1 8B Instruct8B | GPTQ INT4 | 4.97 GB | +1 GB |
| Shot Reader 8B8B | Q3_K_M | 4.93 GB | +1.1 GB |
| Qwen3 8B Instruct8B | AWQ INT4 | 4.9 GB | +1.1 GB |
| Llama 3.1 8B Instruct8B | AWQ INT4 | 4.75 GB | +1.3 GB |
| Aya 23 8B8B | AWQ INT4 | 4.75 GB | +1.3 GB |
| DeepSeek-R1-Distill-Llama-8B8B | AWQ INT4 | 4.75 GB | +1.3 GB |
| InternLM2 7B Chat7B | AWQ INT4 | 4.59 GB | +1.4 GB |
| Qwen2.5 7B Instruct7B | EXL2 4.65bpw | 4.87 GB | +1.1 GB |
| Qwen2.5-Coder 7B Instruct7B | EXL2 4.65bpw | 4.87 GB | +1.1 GB |
| DeepSeek-R1-Distill-Qwen-7B7B | EXL2 4.65bpw | 4.87 GB | +1.1 GB |
| WizardLM-2 7B7B | Q4_K_M | 5.14 GB | +0.9 GB |
| Mistral 7B Instruct v0.37B | Q4_K_M | 5.14 GB | +0.9 GB |
| Zephyr 7B Beta7B | Q4_K_M | 5.14 GB | +0.9 GB |
| K2 Horizon 3.7B3.7B | Q4_K_M | 4.02 GB | +2 GB |
| Gemma 4 E2B ITE2B | Q8_0 | 5.2 GB | +0.8 GB |
| Qwen3.5 4B4B | Q8_0 | 4.97 GB | +1 GB |
| Gemma 3 4B IT4B | Q8_0 | 5.05 GB | +1 GB |
| Spark-X2.6 4B4B | Q8_0 | 4.84 GB | +1.2 GB |
| Qwen3 4B Instruct4B | Q6_K | 4.05 GB | +2 GB |
<4B · 22
| Model | Quant | Est. VRAM | Headroom |
|---|---|---|---|
| Phi-4 Mini Instruct3.8B | Q8_0 | 4.81 GB | +1.2 GB |
| Phi-3.5 Mini Instruct3.8B | Q4_K_M | 4.07 GB | +1.9 GB |
| Granite 4.2 3B3B | Q8_0 | 4.41 GB | +1.6 GB |
| Ministral 3 3B Instruct3B | Q8_0 | 4.26 GB | +1.7 GB |
| Llama 3.2 3B Instruct3B | Q8_0 | 4.05 GB | +2 GB |
| Qwen2.5 3B Instruct3B | Q8_0 | 3.59 GB | +2.4 GB |
| SmolLM3 3B3B | Q8_0 | 3.73 GB | +2.3 GB |
| Metis-1.6 Praxis3B MoE | Q8_0 | 3.4 GB | +2.6 GB |
| Gemma 2 2B Instruct2B | Q8_0 | 3.34 GB | +2.7 GB |
| Qwen3.5 2B2B | Q8_0 | 2.23 GB | +3.8 GB |
| LFM3 1.7B1.7B | Q8_0 | 1.99 GB | +4 GB |
| LFM3-VL 2B2B | Q8_0 | 1.99 GB | +4 GB |
| Qwen3 1.7B Instruct1.7B | Q8_0 | 2.39 GB | +3.6 GB |
| Qwen2.5 1.5B Instruct1.5B | Q8_0 | 1.83 GB | +4.2 GB |
| Llama 3.2 1B Instruct1B | Q8_0 | 1.51 GB | +4.5 GB |
| LFM2.5 1.2B Instruct1.2B | Q8_0 | 1.35 GB | +4.7 GB |
| K2 Horizon 0.9B0.9B | Q8_0 | 1.44 GB | +4.6 GB |
| Qwen3.5 0.8B0.8B | Q8_0 | 0.94 GB | +5.1 GB |
| BotPsy 0.8B0.75B | Q8_0 | 0.9 GB | +5.1 GB |
| Qwen2.5 0.5B Instruct0.5B | Q8_0 | 0.6 GB | +5.4 GB |
| MNN-Pocket 0.3B Instruct0.36B | Q8_0 | 0.46 GB | +5.5 GB |
| MNN-Pocket 0.3B QAT0.30B | Q4_1 | 0.26 GB | +5.7 GB |
How this list is built
Each row is the lowest-perplexity-loss quant of that model whose estimated total — weights plus KV cache at 4K context plus activation buffer — uses at most 88% of the card. That is the calculator's "green" threshold, so each estimate leaves headroom. Raise the context length and the list shortens; the calculator lets you check any combination directly.
A further 8 models load but with no headroom to spare (up to 105% of VRAM) — the Quant Hub’s GPU chips count those too, which is why its number is higher.
Measured on this card
No inference runs were measured on this laptop GPU. Fit figures are estimates, not measured results.
Cards with the same budget
What fits is decided by memory, so every 6GB card of this type returns the same list. These pages are not different answers — they differ in throughput, which this index does not measure per card.
Stepping up
A RTX 5060 Ti 8G (8GB) fits 16 more of the indexed models than this card. RTX 5060 Ti 8G →
Common questions
What is the best local LLM for an RTX 3060 Laptop GPU?
For everyday use, Qwen2.5 3B Instruct at Q8_0 — about 3.6 GB of the card's 6 GB at 4K context, leaving 2.4 GB for a longer window. If you want the largest thing that will load, that is GLM-4-9B-Chat at AWQ INT4 (4.9 GB, 1.1 GB spare). "Best" here means best fit for the memory budget — this index does not run task benchmarks, so it cannot tell you which model is smarter.
Can an RTX 3060 Laptop GPU run K2 Horizon 7B?
No. Its smallest build here, Q4_K_M, needs about 6.5 GB and this card has 6.0 GB usable — short by 0.5 GB before any context beyond 4K. The largest model this card does clear is GLM-4-9B-Chat.
RTX 3060 Laptop GPU or RTX 4050 Laptop GPU for local LLMs?
They hold the same models: 6 GB against 6 GB fits 46 and 46 of 151 respectively at 4K. Bandwidth is not recorded for both cards here, so this page will not rank them on speed.
46 of 151 indexed models fit comfortably in 6GB at 4K context, each at the highest-quality quant that still leaves headroom.
This page's figures change when the model or the runtime does.Last updated 2026-11-22 RSS → /feed.xml