Llama Guard vs LLM Guard
Pick two to four business tools to see their facts side by side.
Llama GuardMeta's open safety models that classify LLM prompts and responses, part of the Purple Llama project.
Remove |
LLM GuardOpen-source scanner toolkit for LLM inputs and outputs; the repository is now archived.
Remove |
|
|---|---|---|
| Category | LLM guardrails and app protection | LLM guardrails and app protection |
| Company | Meta | Protect AI |
| Headquarters | United States | Not stated |
| Deployment | Self-hosted | Self-hosted |
| Sold to | Small business, Mid-size, Enterprise | Small business, Mid-size, Enterprise |
| Platforms | Self-hosted, Linux | Self-hosted, Linux, macOS, Windows |
| Pricing model | Flat fee | Flat fee |
| Starting price | Free (open weights, Llama community licence) | Free (open source) |
| Free tier | Yes | Yes |
| Free trial | No | No |
| Open source | No | Yes (MIT) |
| Single sign-on (SAML) | No | No |
| User provisioning (SCIM) | No | No |
| Certifications | Not stated | Not stated |
| EU data residency | Not stated | Not stated |
| NIST CSF functions | Protect | Protect |
| Verdict | Not stated | Not stated |
| Updated | 4 Nov 2026 | 4 Nov 2026 |
| Links | Website Pricing page All facts | Website Pricing page All facts |
Facts come from each vendor’s documentation, pricing page and audit reports, as of the date shown. Prices change often: check the vendor’s page before you buy.

