GPT-oss-safeguard: Now on ShareAI
ShareAI is committed to bringing you the latest and most powerful AI models—and we’re doing it again today.
OpenAI just released gpt-oss-safeguard, a pair of open-weight safety reasoning models—and you can start using them on ShareAI right now!
The newly released models—gpt-oss-safeguard-20b and gpt-oss-safeguard-120b—interpret your policy at inference time, classify messages/completions/chats accordingly, and show their reasoning for transparent, auditable decisions (Apache 2.0).
🚀 Get started with gpt-oss-safeguard on ShareAI
- Open Playground
- Browse Models
- Docs: Getting Started
- Sign in / Create your account
- Releases
- OpenAI announcement
✨ gpt-oss-safeguard highlights
- Policy-at-inference: Bring your own policy text; update rules anytime without retraining.
- Explainability: See a concise rationale / chain-of-thought for each decision.
- Flexible & domain-aware: Tailor to nuanced domains (forums, reviews, community platforms).
- Open licensing: Apache 2.0 for commercial use, modification, and deployment.
📌 Models available now on ShareAI
gpt-oss-safeguard-20b
Great for lower-latency, cost-sensitive pipelines and quick policy iteration.
gpt-oss-safeguard-120b
Best for deeper reasoning and longer, more complex policies or content.
🛠️ Try gpt-oss-safeguard today
Stay tuned—ShareAI continues to expand our model library so you can build safer, faster, and with full transparency. Happy prompting! 🚀