What LLMs fit in 16GB VRAM? (Oct 2026)
21 models on our list fit in 16 GB of graphics memory. The largest by parameter count is Recursive Flint at UD-Q2_K_XL, 14.63 GB (measured weights; total is an estimate). We keep 0.6 GB free on one graphics card, so 15.4 GB is usable. Each total adds conversation memory for 8,192 tokens (or the model's own maximum, when that is shorter) and 0.8 GB for the software. The total is an estimate.
Updated
Own a 16 GB GPU? Get a 1-line alert when a new top model fits 15.4 GB usable VRAM
21 models on our catalog fit 16 GB today. Get our weekly 1-line alert when a new open-weight LLM lands inside 15.4 GB at Q4_K_M or when 16 GB+ GPU prices drop >10% — zero spam.
Models that fit
Ranked by parameter count, largest first. The best quant is the highest-quality option that fits in 15.4 GB. A measured download is used when we have one. Otherwise the figure is Q4, Q5, Q8 or FP16 from our estimate. The gigabytes are the weights plus conversation memory for 8,192 tokens, or the model's shorter maximum, plus 0.8 GB for the software. The total is an estimate.
| Rank | AI model | Parameters | Best quant | GB | Label |
|---|---|---|---|---|---|
| 1 | Can I run Recursive Flint? | 39.79B, 2.5B active | UD-Q2_K_XL | 14.63 GB | measured weights; total is an estimate |
| 2 | Can I run Nanbeige4.3-33B? | 33.12B (estimate) | IQ2_M | 14.73 GB | measured weights; total is an estimate |
| 3 | Can I run Qwen3 32B? | 32.76B (estimate) | UD-IQ2_XXS | 12.23 GB | measured weights; total is an estimate |
| 4 | Can I run Gemma 4 31B? | 31.27B (estimate) | UD-IQ3_XXS | 14.15 GB | measured weights; total is an estimate |
| 5 | Can I run Qwen3 Coder 30B A3B? | 30.53B (estimate), 3.3B active | UD-Q2_K_XL | 13.40 GB | measured weights; total is an estimate |
| 6 | Can I run Qwen3.8-27B? | 27.78B (estimate) | UD-IQ3_XXS | 12.27 GB | measured weights; total is an estimate |
| 7 | Can I run Gemma 3 27B? | 27.43B (estimate) | Q3_K_M | 15.35 GB | measured weights; total is an estimate |
| 8 | Can I run OrcaCyber Nano 1.0? | 27B | Mix-13G | 14.33 GB | measured weights; total is an estimate |
| 9 | Can I run Gemma 4 26B A4B? | 25.2B (estimate), 3.8B active | UD-IQ4_XS | 14.78 GB | measured weights; total is an estimate |
| 10 | Can I run Mistral Small 3.2 24B? | 24.01B (estimate) | IQ4_XS | 14.90 GB | measured weights; total is an estimate |
| 11 | Can I run gpt-oss-20b? | 21B (estimate), 3.6B active | F16 | 14.79 GB | measured weights; total is an estimate |
| 12 | Can I run DeepSeek R1 Distill Qwen 14B? | 14.77B | Q6_K | 14.53 GB | measured weights; total is an estimate |
| 13 | Can I run Phi-4 14B? | 14.7B (estimate) | Q5 (~0.71 bytes/param) | 12.91 GB | estimate |
| 14 | Can I run Qwen2.5 14B? | 14.7B (estimate) | Q5 (~0.71 bytes/param) | 12.85 GB | estimate |
| 15 | Can I run Gemma 2 9B? | 9.24B (estimate) | Q8 | 12.80 GB | estimate |
| 16 | Can I run Llama 3.1 8B? | 8.03B (estimate) | Q8 | 10.47 GB | estimate |
| 17 | Can I run NVIDIA Nemotron 3.5 Bolt E4B? | 8B (estimate) | UD-Q8_K_XL | 9.26 GB | measured weights; total is an estimate |
| 18 | Can I run Mistral 7B v0.3? | 7.25B (estimate) | Q8 | 9.63 GB | estimate |
| 19 | Can I run Spark-X2.6-4B? | 4.11B (estimate) | BF16 | 9.39 GB | measured weights; total is an estimate |
| 20 | Can I run Llama 3.2 3B? | 3.21B (estimate) | FP16 | 8.16 GB | estimate |
| 21 | Can I run LFM3-1.7B? | 1.73B (estimate) | BF16 | 4.38 GB | measured weights; total is an estimate |
Cards and Macs with 16 GB
The euro price is the figure we store, with the date it was written down. A Mac's gigabytes are shared memory, not graphics memory, so a Mac does not hold the same models as a 16 GB card.
| Machine | Memory | EU price | Date | Label |
|---|---|---|---|---|
| Radeon RX 9060 XT 16GB | 16 GB graphics memory | €564 | New (cheapest in-stock board on GPU PRIX's German board list) | |
| GeForce RTX 4060 Ti 16GB | 16 GB graphics memory | €899 | New price, the lowest in stock in the EU | |
| GeForce RTX 5080 16GB | 16 GB graphics memory | €1,358 | New price, the lowest in stock in the EU | |
| Mac mini M4 16GB | 16 GB shared memory | €549 | price unverified | |
| Mac mini M6 16GB | 16 GB shared memory | €1,069 | New |
Running n8n agents on a local model: n8n local LLM hardware requirements. A 16 GB card is the entry tier for a tool-calling agent. n8n, PostgreSQL and Qdrant need no GPU.
Memory totals are our estimate, including when the weight file was measured. Prices are the euro figures we store, with the date they were written down. How we calculate
Common questions
- What is the largest model that fits in 16 GB of VRAM?
- Recursive Flint is the largest by parameter count, at UD-Q2_K_XL, 14.63 GB (measured weights; total is an estimate). 21 models on our list fit.
- What does the best quant mean for 16 GB?
- The best quant is the highest-quality option that still fits. A measured download is used when we have one at that quality. Otherwise the figure is Q4, Q5, Q8 or FP16 from our estimate. A higher bit count wins. A file that runs only on a Mac is not counted. The gigabytes add conversation memory and the software overhead, so the total is an estimate even when the weight file was measured.
- Which cards and Macs have 16 GB, and what do they cost in the EU?
- Radeon RX 9060 XT 16GB (16 GB graphics memory): €564, New (cheapest in-stock board on GPU PRIX's German board list), 17 Oct 2026. GeForce RTX 4060 Ti 16GB (16 GB graphics memory): €899, New price, the lowest in stock in the EU, 16 Oct 2026. GeForce RTX 5080 16GB (16 GB graphics memory): €1,358, New price, the lowest in stock in the EU, 21 Oct 2026. Mac mini M4 16GB (16 GB shared memory): €549, price unverified, 16 Oct 2026. Mac mini M6 16GB (16 GB shared memory): €1,069, New, 16 Oct 2026.
- How much of 16 GB can a model use?
- Our estimate keeps 0.6 GB free on one graphics card, so 15.4 GB is usable. A Mac with 16 GB of shared memory does not give the graphics all of it. The model table is for one graphics card.