gpt-oss-safeguard-20b
OpenAI's gpt-oss-safeguard-20b is a 21-billion parameter Mixture-of-Experts model designed specifically for safety-oriented tasks. It specializes in content classification and filtering, offering reasoning capabilities and tool-calling functionality for improved performance in trust and safety workflows.
gpt-oss-safeguard-20b specs and pricing
- Provider: OpenAI
- Released: October 29, 2025
- Context window: 131K tokens
- Max output: 66K tokens
- Input: text
- Output: text
- Price: $0.075 input / $0.30 output per 1M tokens
- Reasoning: yes
- Tool calling: yes
What gpt-oss-safeguard-20b is good at
- Automated content moderation
- Filtering large volumes of text
- Building trust and safety classification systems
To-dos to try with gpt-oss-safeguard-20b
- Analyze this dataset of user comments and flag any that violate community guidelines.
- Review this text output and determine if it contains prohibited topics based on my safety policy.
- Categorize these incoming support tickets by severity and potential policy risk.
FAQ
- What is the primary purpose of gpt-oss-safeguard-20b? It is a safety reasoning model built to handle tasks like content classification and LLM filtering.
- Can this model process files or images? No, this model is limited to text input and text output.
- How much does gpt-oss-safeguard-20b cost? gpt-oss-safeguard-20b costs $0.075 per million input tokens and $0.30 per million output tokens through OpenRouter. In todo.is a typical agent step costs about 0.25 credits ($1 = 100 credits).
- What is the context window of gpt-oss-safeguard-20b? gpt-oss-safeguard-20b can read up to 131K tokens at once and write up to 66K tokens in one reply.
- Can I use gpt-oss-safeguard-20b without coding? Yes. In todo.is you pick gpt-oss-safeguard-20b as your agent's model and write to-dos in plain words, on the web, the Android app, WhatsApp or Telegram. No API key needed: credits are included, or you can bring your own key.
JavaScript is required to use the todo.is app.