GPT OSS Safeguard 20B

openai

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 and safety labeling.

Try Now

Capabilities

Extended Thinking

Tool Use

Example Use Cases

Content classification and moderation

LLM output filtering

Trust and safety labeling

Technical Specifications

Context Window

131,072 tokens

Max Output

65,536 tokens

Pricing

Token Costs (per 1M tokens)

Cache Miss Input

$0.075

Non-Reasoning Output

$0.30

Cache Read Input

$0.037

Tool Costs (per 1K calls)

Web Search

$15

Code Execution

$0.19

⚠️ Legacy

Made legacy on

Reason

Not a chat model

Recommended Replacement

Qwen3.5 Plus