Catalog

All models

Every model for which we have at least a price or a specification from an open source. An unfiltered list: it includes deprecated models and models without a single independent score.

Model Developer Input, $ per 1M Output, $ per 1M Context Released
GLiGuard LLM Guardrails 300M Fastino Labs $0.15 $0.15 8K Apr 2026
GPT OSS Safeguard 120B OpenAI $0.15 $0.6 131K Oct 2025
GPT OSS Safeguard 20B OpenAI $0.07 $0.2 131K Oct 2025
Granite Guardian 3.2 2B IBM $0.1 $0.1 8K —
Granite Guardian 3.3 8B IBM $0.2 $0.2 8K —
Llama 3.1 Nemotron Safety Guard 8B V3 Meta AI — — 128K Oct 2025
Llama Guard 2 8B Meta AI $0.2 $0.2 8K —
Llama Guard 3 11B Vision Meta AI $0.35 $0.35 128K —
Llama Guard 3 1B Meta AI $0.1 $0.1 131K —
Llama Guard 3 8B Meta AI $0.02 $0.06 131K Jan 2025
Llama Guard 4 12B Meta AI $0.18 $0.18 164K Apr 2025
Llamaguard 7B Meta AI $0.2 $0.2 4K —
Meta Llama Guard 4 12B Meta AI $0.21 $0.21 131K Jan 2025
Meta Llama Prompt Guard 2 22M Meta AI $0.01 $0.01 1K Oct 2024
Meta Llama Prompt Guard 2 86M Meta AI $0.01 $0.01 1K Oct 2024
Omni Moderation 2024-09-26 OpenAI — — 33K —
Omni Moderation Latest OpenAI — — 33K —
Qwen3Guard Gen 0.6B Alibaba (Qwen) — — 33K Jan 2026
Qwen3Guard Gen 8B Alibaba (Qwen) — — 33K Jan 2026
Text Moderation 007 OpenAI — — 33K —
Text Moderation Latest OpenAI — — 33K —
Text Moderation Stable OpenAI — — 33K —

The price is the developer's own, if they sell this model; if not, the cheapest offer among providers is taken, with the input-and-output pair coming entirely from one price list rather than as two minimums from different ones. Base per-token tier only. A dash means the price is absent from open sources, not that the model is free.