---
id: "openai/gpt-oss-safeguard-20b"
also: ["openai/gpt-oss-safeguard-20b", "hf/openai/gpt-oss-safeguard-20b", "vercel/openai/gpt-oss-safeguard-20b"]
cheapest_list_usd_per_m: "0.07 in, 0.2 out, at bedrock on vercel"
cheapest_effective_usd_per_m: "0.0594 in, 0.3 out, at groq on vercel"
context: 131072
max_output: 65536
input: ["text"]
output: ["text"]
released: "2025-10-29"
api: ["/api/v1/chat/completions", "/api/v1/completions", "/api/v1/messages", "/api/v1/responses"]
params: ["include_reasoning", "max_tokens", "reasoning", "response_format", "seed", "stop", "structured_outputs", "temperature", "tool_choice", "tools", "top_p"]
endpoints: 4
gateways: "vercel (2), hf (1), openrouter (1)"
no_training: 3
trains: 0
retains_prompts: 0
---
# OpenAI: gpt-oss-safeguard-20b

OpenAI's first open weight reasoning model specifically trained for safety classification tasks. Fine-tuned from GPT-OSS, this model helps classify text content based on customizable policies, enabling bring-your-own-policy Trust & Safety AI where your own taxonomy, definitions, and thresholds guide classification decisions.