What changed
OpenAI released o3 and o4-mini with the ability to use web search, Python, file analysis, visual inputs, and image manipulation within their reasoning process in ChatGPT and the API.
Why it matters
The model could decide when and how to gather evidence or compute an answer, making tool orchestration part of reasoning rather than a separate prompt pattern.
What to do
Give each tool a scoped permission, schema validator, cost limit, and deterministic postcondition; review the trace when a result matters.
Caveat
Autonomous tool selection can compound errors or spend, and benchmark leadership was vendor-reported. Models still hallucinate and can misuse visual evidence.