Models · OpenAI
Pluely × OpenAI: gpt-oss-safeguard-20b
Text → TextNo setup · no API key
gpt-oss-safeguard-20b is a safety reasoning model from OpenAI built upon gpt-oss-20b. This open-weight, 21B-parameter Mixture-of-Experts (MoE) model offers lower latency for safety tasks like content classification, LLM filtering, and trust...
How OpenAI: gpt-oss-safeguard-20b works in Pluely
- One tap to switch. Open the overlay's model picker, search "OpenAI: gpt-oss-safeguard-20b", click — your next question uses it. See the model picker.
- Works everywhere in the app. Ask mode questions, automatic responses in meetings, transcript summaries, and follow-ups all run on whichever model you selected.
- Text-first. OpenAI: gpt-oss-safeguard-20b is a text model — Pluely automatically disables the screenshot controls while it's selected so you never send an image a model can't read.
- Metered simply. Each completion counts as one AI request on your monthly plan — see plans & usage.
- Private by design. Conversations stay on your device, and Pluely never exposes its serving infrastructure to the app.