About

Why use this instead of the official site?

The official GSoC website lists organizations but links out to external pages for project ideas — scattered across Google Docs, GitHub wikis, GitLab issues, and various websites. This makes it hard to search, compare, and explore what's available.

GSoC 2026 Explorer solves this by:

• Full-text search across everything — press Cmd+K (or Ctrl+K) to search across all organization names, descriptions, technologies, topics, and project ideas content at once
• Advanced combined filters — filter by technology and topic tags simultaneously, with shareable URLs (e.g. ?tech=python&topic=web)
• All ideas in one place — every ideas page has been scraped and rendered in a consistent format, with table of contents for easy navigation
• LLM-ready structured data — the repository contains all data in JSON and Markdown, ready to be fed to language models for analyzing opportunities, matching skills, or generating summaries
• Fast and responsive — static site with instant filtering, dark mode, and no page reloads

Data overview

• 185 organizations
• 180 ideas pages successfully scraped
• 261 unique technologies
• 510 unique topics

How the data was collected

A Python scraper queries the GSoC public API to get all participating organizations, then fetches each organization's ideas page using multiple strategies:

• GitHub blob/wiki/gist URLs converted to raw URLs for clean Markdown
• Google Docs exported as HTML or plain text
• GitHub and GitLab issues fetched via CLI/API
• HTML pages extracted via trafilatura or markdownify
• Playwright headless browser for JavaScript-rendered pages

Source code

Both the scraper and this website are open source. The repository also contains the raw scraped data in Markdown and JSON format.