Google’s Deep Research is an agentic AI feature that autonomously plans multi-step research and browses the web, reasons iteratively over findings, and generates detailed, cited reports. It takes a few minutes but creates detailed reports.
It is now available through an API so it can be built into custom applications. The prompt can take formatting instructions and follow up questions can be made.
Docs: https://ai.google.dev/gemini-api/docs/deep-research
Announcement: https://blog.google/technology/developers/deep-research-agent-gemini-api/
Example just using curl in the terminal:
# 1. Start the research task
curl -X POST "https://generativelanguage.googleapis.com/v1beta/interactions" \
-H "Content-Type: application/json" \
-H "x-goog-api-key: $GEMINI_API_KEY" \
-d '{
"input": "Research the history of Google TPUs.",
"agent": "deep-research-pro-preview-12-2025",
"background": true
}'
This will return an ID which you can poll for the result.
# 2. Poll for results (Replace INTERACTION_ID)
curl -X GET "https://generativelanguage.googleapis.com/v1beta/interactions/INTERACTION_ID" \
-H "x-goog-api-key: $GEMINI_API_KEY"I asked it to “Investigate use cases and examples of Claude Skills.”
The response was a 2,800 word report with citations. Check it out here:
https://github.com/wanacode/Claude-Skills-Report
Cost came to US$1.35