What can you run on a 16GB graphics card?
With 16GB VRAM you can run 13 of the 51 open-source models we track — up to about 12B parameters at Q4 quantization. The most popular that fits is gemma-4-12B-it, needing 9.6 GB and generating around ~65 tok/s on typical hardware of this size.
Typical machines: RTX 4080 · RTX 5080 · RTX 4060 Ti 16GB · RX 7800 XT · Arc A770. After the operating system takes its share, about 16 GB is free for a model.
Having the memory is not the same as having the speed
16GB VRAM decides whether a model loads. How fast it answers is decided by memory bandwidth — how quickly your machine can read the model, which it must do once for every word it writes. Machines with 16GB VRAM span 288–960 GB/s, so the fastest is about 3.3× quicker than the slowest with identical capacity. The table quotes a RTX 4080; pick yours below to see the difference. Why bandwidth and not the processor →
Models that run on 16GB VRAM in 2026
| Model↕ | Size↕ | Memory↑ | Est. speed↕ | Downloads↕ | Feels like |
|---|---|---|---|---|---|
| LFM2.5-230MLiquid Ai | 230M | 808 MBRuns well | ~3038 tok/s | 58,011 | Fast |
| MiniCPM5-1BOpenBMB | 1.1B | 1.3 GBRuns well | ~714 tok/s | 667,154 | Fast |
| MiniCPM5-2BOpenBMB | 2.5B | 2.3 GBRuns well | ~298 tok/s | 206,774 | Fast |
| LFM2.5-2.6BLiquid Ai | 2.7B | 2.5 GBRuns well | ~278 tok/s | 98,713 | Fast |
| LFM2.5-VL-3BLiquid Ai | 3.1B | 2.5 GBRuns well | ~278 tok/s | 31,296 | Fast |
| North-Micro-Vision-InstructCohere | 2.5B | 2.5 GBRuns well | ~311 tok/s | 120,354 | Fast |
| granite-4.2-3bIbm | 3.7B | 3.1 GBRuns well | ~208 tok/s | 32,017 | Fast |
| Mage-VLMicrosoft | 4.7B | 3.9 GBRuns well | ~172 tok/s | 29,690 | Fast |
| LFM2.5-8B-A1BLiquid Ai · MoE | 8.5B | 5.9 GBRuns well | ~765 tok/s1B active | 80,632 | Fast |
| Ling-3.0-tinyInclusionai · MoE | 7.9B | 6.2 GBRuns well | ~483 tok/s1.6B active | 25,727 | Fast |
| granite-4.2-8bIbm | 8.8B | 6.6 GBRuns well | ~87 tok/s | 52,946 | Fast |
| Ornith-1.5-9BDeepreinforce | 9.7B | 6.9 GBRuns well | ~81 tok/s | 453,339 | Fast |
| gemma-4-12B-itGoogle DeepMind | 12B | 9.6 GBRuns well | ~65 tok/s | 2,958,440 | Fast |
Sorted by memory, smallest first — click any column to change it.
Step up to 24GB VRAM and 7 more models come within reach
Including Qwen3.8-27B (28B), diffusiongemma-26B-A4B-it (26B) and Muse-Glimmer-30B (30B) — around ~40 tok/s on a typical machine that size.
See everything 24GB VRAM runs →All figures at Q4_K_M quantization, 4K context. “Memory” includes the model, its context cache and runtime overhead. Speed is calculated, not benchmarked: 717 GB/s × 0.65 ÷ bytes read per token, assuming a RTX 4080. Scale it by your own machine's bandwidth from the list above. Mixture-of-experts models read only their active parameters per token, which is why some large ones outrun smaller dense models.
What you'd need more memory for
The next models up, and what they ask for.