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...
Developer
OpenAI
Release Date
October 29, 2025
Context Window
131,072 tokens≈ 175 words
API Access
Publicly AvailableIntegrate via official API
Input Cost
$0.07per million tokens
Output Cost
$0.30per million tokens
SHARE MODEL:
Vetted Benchmarks
No recorded benchmark scores available for this model.
How does this model compare?
Evaluate benchmark standing and performance rankings vs all other tracked models.