Llama Guard or ShieldGemma: which safety filter to pick
Both models do the same thing: check user questions and bot replies for dangerous content against a list of categories, both are in Ollama and both run on a CPU. Neither officially supports Russian, so on a Russian-language project the filter has to be tested on your own examples rather than trusted blindly. Llama Guard spans 1B to 12B, version 4 from 2025-04 also checks images, and the Llama family licenses allow commercial use with restrictions. ShieldGemma runs from 2B to 27B, ShieldGemma 2 from 2025-03 checks generated images before publishing, and the Gemma Terms of Use likewise allow commercial use with restrictions on types of use. If you already build on one of these families, take the filter from the same one.
Comparison based on catalog data
| Parameter | Llama Guard | ShieldGemma |
|---|---|---|
| Category | Moderation and safety | Moderation and safety |
| Developer | Meta, USA | Google, USA |
| Releases | Dec 2023 – Apr 2025 | Jul 2024 – Mar 2025 |
| Sizes | 1B – 12B | 2B – 27B |
| Hardware | Laptop, 1 GPU | Laptop, 1 GPU |
| Commercial use | Commercial use with conditions | Commercial use with conditions |
| License | Llama 2, 3.1 and 4 licenses: commercial use allowed with restrictions | Gemma Terms of Use: commercial use allowed with restrictions on types of use |
| Russian | Not supported | Not supported |
| Ollama | Yes | Yes |
| Without GPU | Yes | Yes |
| Tasks |
|
|
Choose Llama Guard if
- You need a very small filter: Llama Guard comes in 1B
- You already run Llama models and want one family
- You need a report naming the exact rule category that was violated
Choose ShieldGemma if
- You need a large version for demanding moderation: ShieldGemma goes to 27B
- Your main job is checking generated images before publishing
- You are already building on Gemma models
Other comparisons
- Mistral or Llama: which to choose for business
- Mistral or Qwen: which to choose for business
- Qwen or GigaChat: which to choose for business
- Qwen or Llama: which to choose for business
- GigaChat or YandexGPT: which to choose for business
- DeepSeek or Qwen: which to choose for business
- Gemma or Llama: which to choose for business
- Whisper or GigaAM: which to choose for business
- Whisper or Parakeet: which to choose for business
- Silero or Piper: which to choose for business
- FLUX or Stable Diffusion: which to choose for business
- FLUX or Qwen-Image: which to choose for business
Need a model for your task?
An open model can run on your own server: data stays in-house, there is no per-request fee, and the model can be fine-tuned on your documents.
- SelectThe model and size for your task and hardware budget
- DeployOn your server or in a closed network, with an API
- Fine-tuneOn your data, or connect a knowledge base
- IntegrateInto your CRM, ERP, bot, website or team chat