NVIDIA · 2023-07
Best local LLMs for the RTX 4060 Ti 16 GB
16 GB VRAM · 288 GB/s · ~15 GB usable. Cheapest way to 16 GB of VRAM — bandwidth is low, so it’s about fitting, not speed.
Check RTX 4060 Ti 16 GB price →Try it with your context & use case
Preset to the RTX 4060 Ti 16 GB. Change the context length or filter by use case.
16 models run well and 7 run tight on RTX 4060 Ti 16 GB (any context — just fitting the weights) — 15 GB usable.
| Model | Size | Fit | Best quant | Needs | Memory | Speed |
|---|---|---|---|---|---|---|
| Mistral Small 3 24B | 23.6B | Runs well | Q4_K_M | 15 GB | ~15 tok/s | |
| Qwen3 14B | 14.8B | Runs well | Q6_K | 13 GB | ~17 tok/s | |
| Qwen2.5 14B | 14.8B | Runs well | Q6_K | 13 GB | ~17 tok/s | |
| DeepSeek-R1 Distill Qwen 14B | 14.8B | Runs well | Q6_K | 13 GB | ~17 tok/s | |
| Phi-4 14B | 14.7B | Runs well | Q6_K | 12 GB | ~17 tok/s | |
| Mistral Nemo 12B | 12.2B | Runs well | Q8_0 | 13 GB | ~16 tok/s | |
| Gemma 3 12B | 12.2B | Runs well | Q8_0 | 13 GB | ~16 tok/s | |
| Gemma 2 9B | 9.2B | Runs well | Q8_0 | 10 GB | ~21 tok/s | |
| Qwen3 8B | 8.2B | Runs well | Q8_0 | 9.2 GB | ~24 tok/s | |
| Llama 3.1 8B | 8.0B | Runs well | Q8_0 | 9.0 GB | ~24 tok/s | |
| Qwen2.5-Coder 7B | 7.6B | Runs well | Q8_0 | 8.6 GB | ~26 tok/s | |
| Mistral 7B v0.3 | 7.3B | Runs well | FP16 | 15 GB | ~14 tok/s | |
| Gemma 3 4B | 4.3B | Runs well | FP16 | 9.1 GB | ~24 tok/s | |
| Qwen3 4B | 4B | Runs well | FP16 | 8.5 GB | ~26 tok/s | |
| Llama 3.2 3B | 3.2B | Runs well | FP16 | 7.0 GB | ~32 tok/s | |
| Llama 3.2 1B | 1.2B | Runs well | FP16 | 3.2 GB | ~84 tok/s | |
| Qwen3 32B | 32.8B | Runs (tight) | Q2_K | 14 GB | ~15 tok/s | |
| Qwen2.5 32B | 32.8B | Runs (tight) | Q2_K | 14 GB | ~15 tok/s | |
| Qwen2.5-Coder 32B | 32.8B | Runs (tight) | Q2_K | 14 GB | ~15 tok/s | |
| DeepSeek-R1 Distill Qwen 32B | 32.8B | Runs (tight) | Q2_K | 14 GB | ~15 tok/s | |
| Qwen3 30B-A3B (MoE)MoE | 30.5B | Runs (tight) | Q3_K_M | 15 GB | ~84 tok/s | |
| Gemma 3 27B | 27.4B | Runs (tight) | Q3_K_M | 14 GB | ~15 tok/s | |
| Gemma 2 27B | 27.2B | Runs (tight) | Q3_K_M | 14 GB | ~16 tok/s | |
| DeepSeek-R1 671B-A37B (MoE)MoE | 671B | Won't fit | — | 273 GB | — | |
| Qwen3 235B-A22B (MoE)MoE | 235B | Won't fit | — | 96 GB | — | |
| Llama 4 Scout (MoE)MoE | 109B | Won't fit | — | 45 GB | — | |
| Qwen2.5 72B | 72.7B | Won't fit | — | 30 GB | — | |
| Llama 3.3 70B | 70.6B | Won't fit | — | 29 GB | — | |
| DeepSeek-R1 Distill Llama 70B | 70.6B | Won't fit | — | 29 GB | — | |
| Mixtral 8x7B (MoE)MoE | 46.7B | Won't fit | — | 20 GB | — |
Estimates, computed in your browser — VRAM, quantisation and speed vary with your runtime and settings. How we estimate →
Top picks for the RTX 4060 Ti 16 GB
Every model on the RTX 4060 Ti 16 GB
| Model | Size | Fit | Best quant | Needs | Speed |
|---|---|---|---|---|---|
| Qwen3 14B | 14.8B | Runs well | Q6 | 14 GB | ~17 tok/s |
| Qwen2.5 14B | 14.8B | Runs well | Q6 | 14 GB | ~17 tok/s |
| DeepSeek-R1 Distill Qwen 14B | 14.8B | Runs well | Q6 | 14 GB | ~17 tok/s |
| Phi-4 14B | 14.7B | Runs well | Q6 | 14 GB | ~17 tok/s |
| Mistral Nemo 12B | 12.2B | Runs well | Q8 | 15 GB | ~16 tok/s |
| Gemma 3 12B | 12.2B | Runs well | Q6 | 13 GB | ~21 tok/s |
| Gemma 2 9B | 9.2B | Runs well | Q8 | 13 GB | ~21 tok/s |
| Qwen3 8B | 8.2B | Runs well | Q8 | 10 GB | ~24 tok/s |
| Llama 3.1 8B | 8.0B | Runs well | Q8 | 10 GB | ~24 tok/s |
| Qwen2.5-Coder 7B | 7.6B | Runs well | Q8 | 9.0 GB | ~26 tok/s |
| Mistral 7B v0.3 | 7.3B | Runs well | Q8 | 9.2 GB | ~27 tok/s |
| Gemma 3 4B | 4.3B | Runs well | FP16 | 10 GB | ~24 tok/s |
| Qwen3 4B | 4B | Runs well | FP16 | 9.6 GB | ~26 tok/s |
| Llama 3.2 3B | 3.2B | Runs well | FP16 | 7.8 GB | ~32 tok/s |
| Llama 3.2 1B | 1.2B | Runs well | FP16 | 3.4 GB | ~84 tok/s |
| Qwen3 30B-A3B (MoE)MoE | 30.5B | Runs (tight) | Q2 | 14 GB | ~98 tok/s |
| Gemma 2 27B | 27.2B | Runs (tight) | Q2 | 15 GB | ~18 tok/s |
| Mistral Small 3 24B | 23.6B | Runs (tight) | Q3 | 13 GB | ~18 tok/s |
| DeepSeek-R1 671B-A37B (MoE)MoE | 671B | Won't fit | — | 303 GB | — |
| Qwen3 235B-A22B (MoE)MoE | 235B | Won't fit | — | 98 GB | — |
| Llama 4 Scout (MoE)MoE | 109B | Won't fit | — | 46 GB | — |
| Qwen2.5 72B | 72.7B | Won't fit | — | 33 GB | — |
| Llama 3.3 70B | 70.6B | Won't fit | — | 32 GB | — |
| DeepSeek-R1 Distill Llama 70B | 70.6B | Won't fit | — | 32 GB | — |
| Mixtral 8x7B (MoE)MoE | 46.7B | Won't fit | — | 21 GB | — |
| Qwen3 32B | 32.8B | Won't fit | — | 16 GB | — |
| Qwen2.5 32B | 32.8B | Won't fit | — | 16 GB | — |
| Qwen2.5-Coder 32B | 32.8B | Won't fit | — | 16 GB | — |
| DeepSeek-R1 Distill Qwen 32B | 32.8B | Won't fit | — | 16 GB | — |
| Gemma 3 27B | 27.4B | Won't fit | — | 16 GB | — |
Similar hardware
FAQ
What is the best LLM for the RTX 4060 Ti 16 GB?
For general use, Qwen3 14B is the strongest model that runs well on the RTX 4060 Ti 16 GB. See the picks-by-use-case below for coding, reasoning and more.
How much can the RTX 4060 Ti 16 GB run?
The RTX 4060 Ti 16 GB has 16 GB of VRAM, of which about 15 GB is usable for a model. That runs 15 of our tracked models well and 3 more at a tight quantisation.
Estimates at 8K context — see how we compute these.