Agentic Researcher
.
model-driven loop: the LLM decides every step
Configuration
OpenAI API key
Sent only to api.openai.com from your browser. Not stored anywhere. Cleared on refresh.
Research goal
Model
gpt-4o-mini (cheap)
gpt-4o
gpt-4.1
Step budget
10
15
25
Deploy agent
Stop
Agent trace
step
0
searches
0
notes
0
idle
Waiting for deployment. The agent will decide its own searches, notes, reflection, and when to finish.
Report
Copy markdown
Download .md
The agent writes here when it decides its research is complete.