Open-weight models for agentic workflows
gpt-oss-120b and gpt-oss-20b are suitable for agentic workflows needing instruction following and tool use.
gpt-oss-120b or gpt-oss-20b
Job
Agentic workflows requiring strong instruction following, tool use (web search, Python code execution), and reasoning.
Reasoning
These models are text-only, compatible with Responses API, and designed for agentic workflows with capabilities like adjustable reasoning effort, full CoT, and structured outputs.
As of
2025-08-05
What it does and when to use it
The gpt-oss-120b and gpt-oss-20b models are recommended for agentic workflows. They offer strong instruction following, tool use, and adjustable reasoning capabilities.
Sources and credit
- gpt-oss-120b & gpt-oss-20b Model Card ↗OpenAI · OpenAI News · Aug 4, 2025 · paraphrased
First seen on Subvolts Aug 4, 2025. Updated Sep 11, 2026. Spotted a problem? Tell us.
Where it shows up
- gpt-oss-120b & gpt-oss-20b Model CardOpenAI · OpenAI News · Aug 4, 2025
Terms used
Related in the PlaybookPlaybook →
The Playbook: reusable how-tos, prompts, configs and tips pulled out of the news.
- ToolHugging Face for Holo4 Model AccessDownload Holo4 LLM variants for agentic tasks. · 15 min
- Safety tipProduct-level safety mitigations for agentsEmploy product-level safety measures like agent sandboxing and configurable network access for AI models. · 15 min
- ToolOpen-weight reasoning models for custom safety classificationOpen-weight models that interpret developer-provided policies for content classification at inference time. · 15 min
- HeuristicUse AI agents to orchestrate other AI agentsHave one AI agent act as an orchestrator, directing other agents and monitoring their progress. · 5 min
- HeuristicEvading AI skill detection by moving payloads to natural languageBypass AI skill detection by embedding malicious payloads within natural language instructions. · 30 min
- HeuristicFraming malicious actions as legitimate skill purposesDisguise malicious actions as the intended function of an AI agent skill. · 30 min
Questions people ask
- Which AI tools does this work with?
- Any current model. It was seen with in the sources.
- Where did this come from?
- OpenAI (OpenAI News). Subvolts collected, normalized and credited it; the link above opens the original.