Entity Extraction with Gemini in BearAudit

BearAudit can run entity extraction on any stored article. It sends the page's title, description, and main text to the Gemini API and gets back structured data: people, places, organizations, events, and a short summary. Results are saved in the app for that article.

What you get

• people — Names mentioned in the text.
• places — Cities, countries, landmarks, etc.
• organizations — Companies, institutions, groups.
• events — Named events (e.g. product launches, conferences).
• summary — A brief 1–3 sentence summary of the content.

Use it for content audits, knowledge graphs, or quickly scanning what an article is about without reading the whole thing.

How to get a Gemini API key

You need a Google API key for Gemini. The simplest way is Google AI Studio (free tier available):

1. Go to Google AI Studio and sign in with your Google account.
2. In the left sidebar, open "Get API key" (or use the key icon in the top bar).
3. Click "Create API key". You can create a key in a new project or pick an existing Google Cloud project.
4. Copy the key — it looks like AIza.... Store it somewhere safe; you won’t see the full key again.

That key is all BearAudit needs. Free tier limits are usually enough for occasional entity extraction; if you hit limits, check Google AI pricing or create a key in Google Cloud with billing enabled.

How to enable it in BearAudit

1. Settings → Integrations — Open the Integrations section and find Gemini. Paste your API key in the field (it’s stored locally in BearAudit only).
2. Alternative: environment variables — You can instead set GEMINI_API_KEY or GOOGLE_API_KEY in your system environment; BearAudit will use that if no key is set in the app.
3. Run extraction — Open a crawled article in BearAudit and use "Extract entities". The app sends the article’s title, description, and main text to Gemini and shows (and can save) people, places, organizations, events, and a summary.

Only the text you choose to send is sent to the API; the rest of your crawl data stays on your machine. Extraction is optional—the rest of BearAudit works without it. The same Gemini integration is used for the Action plan feature, which generates a prioritised SEO action plan from your crawl and analytics data.

When it's useful

• Content inventory — Quickly tag articles by topic, entities, or summary.
• Internal search or recommendations — Use extracted entities to power search or "related content."
• Compliance or fact-checking — See which people, places, and organizations are mentioned across a set of articles.

If you don't need entity extraction, leave the integration blank; crawling, verification, and exports work as usual.