Market research usually takes hours and quickly drains your focus. In this tutorial, you’ll set up Gemini 2.5 Computer Use so it can browse competitor sites, extract the key info, and deliver a clean report in minutes.
Install Python
Install Python on Mac or Windows, then confirm it’s working in Terminal or PowerShell with:
python --versionpython3 --versionSet up project folder
Create a clean folder for this project and launch a terminal inside it so everything stays organized.
Clone repo and install dependencies
Clone the Computer Use GitHub repo, activate the virtual environment, then install packages. Also run the browser setup so automation works:
pip install (repo requirements)playwright install
Get Gemini API key
Create an API key in Google AI Studio, and link billing if needed so the tool can use Gemini 2.5.
Set API key as environment variable
Add your API key as an environment variable instead of pasting it into files, so it stays secure and reusable.
Write clear research query
Be specific about what to collect and how to format the output.
Include:
- competitors or websites
- what to extract (pricing tiers, limits, key features)
- desired output (comparison table + summary + recommendation)
Example query:
Research the AI meeting assistant market. Visit Otter, Fireflies, Avoma, and Fathom. Extract pricing tiers, what each tier includes, user limits, and standout features. Output a comparison table plus a final recommendation for a 10-person sales team.Run agent command
Run:
python main.py --query="YOUR_RESEARCH_QUERY"Gemini will launch Chrome, browse, collect data, and generate a report you can save.
Note: Run two passes. First pass collects facts, second pass compares and recommends using your criteria.