Meta AI · llama
Can I run llama-guard3?
Llama Guard 3 is a series of models fine-tuned for content safety classification of LLM inputs and responses.
8B parameters128K contexttextLLAMA 3.1 COMMUNITY LICENSE AGREEMENT12 builds
llama-guard3 system requirements
Memory needed at 4K context, counting model weights, KV cache, compute buffers and runtime overhead. Download sizes come from the official Ollama registry.
| Build | Quantization | Download | Memory needed |
|---|---|---|---|
| llama-guard3:1b-q4_0 | Q4_0 | 0.9 GB | ~1.7 GB(est.) |
| llama-guard3:1b-q4_K_M | Q4_K_M | 0.9 GB | ~1.7 GB(est.) |
| llama-guard3:1b-q5_K_M | Q5_K_M | 1.0 GB | ~1.9 GB(est.) |
| llama-guard3:1b-q6_K | Q6_K | 1.2 GB | ~2.1 GB(est.) |
| llama-guard3:1b | Q8_0 | 1.5 GB | ~2.5 GB(est.) |
| llama-guard3:1b-fp16 | F16 | 2.8 GB | ~4.4 GB(est.) |
| llama-guard3:8b-q4_0 | Q4_0 | 4.3 GB | ~6.6 GB(est.) |
| llama-guard3:8b | Q4_K_M | 4.6 GB | ~6.9 GB(est.) |
Figures marked (est.) approximate the KV cache because this model's architecture details are not published. They are indicative rather than exact.
How to run llama-guard3 locally
- 1.Install Ollama for Windows, macOS or Linux.
- 2.Run this in a terminal:
ollama run llama-guard3:1b-q4_0 - 3.The weights download on first run, then an interactive prompt opens.
Frequently asked questions about llama-guard3
- How much VRAM do I need to run llama-guard3?
- At 4K context, the smallest published build of llama-guard3 (Q4_0) needs roughly 1.7 GB of GPU memory once weights, KV cache, compute buffers and runtime overhead are counted. Bigger quantizations and longer context windows need more. It can also run partly or entirely on the CPU using system RAM, more slowly.
- Can I run llama-guard3 without a dedicated GPU?
- Yes, but slowly. Without a GPU the model runs on the CPU using system RAM, which usually means a few tokens per second rather than dozens. You would need at least 1.7 GB of free RAM for the smallest build.
- How big is the llama-guard3 download?
- The smallest published build is 0.9 GB. There are 12 builds in total, the largest being 15.0 GB. Leave some extra free disk space beyond the download itself.
- How do I run llama-guard3 locally?
- Install Ollama, then run "ollama run llama-guard3:1b-q4_0" in a terminal. The weights download on first use and an interactive session opens.
Will llama-guard3 run on your PC?
Scan your hardware once and get a verdict for this model and every build — free, no account.
Model data from the official Ollama registry · verified 13/09/2026