Textorum.io is built around a simple principle: your text belongs to you. This page explains exactly what happens when you use each feature — what runs entirely in your browser and what uses external AI services.
Why browser-based processing matters
Most text analysis does not require a server. Running it directly in your browser lowers privacy risk, removes upload latency, and avoids creating copies of your content on infrastructure you do not control. Modern JavaScript engines are fast enough to compute word counts, readability scores and keyword statistics on documents of typical length without noticeable delay.
Textorum.io applies a privacy-first architecture: features that can run locally do run locally, and only features that genuinely require a remote model (AI rewriting, AI metadata generation) make network calls.
Text analysis — 100% local, no server contact
The following features run as local JavaScript inside your browser. They do not intentionally make any network request to Textorum.io or to any third party in order to compute their result:
- Word count
- Character count
- Sentence count
- Paragraph count
- Readability score (Flesch reading ease)
- Flesch–Kincaid grade level
- Keyword density and top keywords
- SEO score
- Reading time estimate
The implementation lives in src/textAnalysis.ts. It is pure JavaScript with no fetch calls, no SDK imports, and no background uploads. Results are computed synchronously as you type.
Note: No text is intentionally transmitted to Textorum.io servers for these features.
| Feature | Runs locally? | Leaves browser? |
|---|---|---|
| Word / character count | Yes | Never intentionally |
| Readability (Flesch) | Yes | Never intentionally |
| Keyword density | Yes | Never intentionally |
| SEO score | Yes | Never intentionally |
| Reading time | Yes | Never intentionally |
What Textorum.io does not do
- No advertising trackers
- No selling of user text
- No permanent storage of analysis text
- No account required for local analysis tools
- No hidden background uploads for analysis tools
- No cross-site tracking
- No marketing cookies
AI features — processed via secure API
Three features genuinely need a remote large language model: AI Rewrite, AI Optimize and the AI SEO Metadata Generator. When you activate one of them, the relevant text is transmitted over HTTPS to a Supabase Edge Function running on Supabase infrastructure. The edge function forwards the request to Google Gemini via the Lovable AI Gateway, then returns the AI response to your browser.
Transmission happens for request fulfillment only. Textorum.io performs no intentional database writes of your text; it is processed transiently in memory while the request is in flight.
Note: Data sent to Google Gemini via the Lovable AI Gateway API is subject to Google's API terms of service. When using enterprise or API-tier access, Google's terms specify that submitted data is not used to train Google's models. Users handling sensitive content should review Google's current API privacy terms to confirm the applicable conditions for their use case. See the Google Gemini API Terms of Service.
What Textorum.io logs (not your text)
Textorum.io logs the following operational metadata for security and reliability — your text content is not intentionally logged:
- Anonymous request ID
- Character count (not the text itself)
- Processing time (latency in milliseconds)
- Language selected
- Mode selected (rewrite / optimize / metadata)
Note: No text content is intentionally logged by Textorum.io infrastructure.
| AI Feature | Destination | Stored by Textorum.io? |
|---|---|---|
| AI Rewrite / Optimize | Supabase → Google Gemini | Not intentionally |
| AI SEO Metadata | Supabase → Google Gemini | Not intentionally |
Note: If you prefer zero intentional data transmission, use only the local analysis tools (word count, readability, keyword density) — these do not intentionally contact any server.
Anonymous authentication
To verify AI requests securely, Textorum.io creates an anonymous session — no email, no name, no password. The session yields a short-lived JWT used by the edge function to validate requests. The anonymous user ID is stored in your browser's localStorage. It is not linked to any personal information or identity, and you can remove it at any time by clearing your browser data. No profile or personal data is collected.
Analytics
Textorum.io plans to integrate Plausible Analytics — a privacy-friendly, cookie-free service that collects only aggregated, anonymous pageview statistics. No text content or personal data is collected by Plausible.
Currently, no tracking or analytics is active on Textorum.io.
Hosting & infrastructure
- Website hosted by Lovable Labs, Stockholm, Sweden
- CDN and DDoS protection via Cloudflare
- HTTPS transport throughout
- AI requests processed via
Supabase Edge Functions
Note: Local analysis tools process text directly in the browser without intentionally routing analysis content through Textorum.io infrastructure.
Data flow overview
Two distinct paths exist, depending on which feature you use:
User types text in browser ↓ Local JavaScript (textAnalysis.ts) ↓ Results computed in browser ↓ No network request ↓ Text stays on device
User activates AI feature ↓ Text sent via HTTPS ↓ Supabase Edge Function ↓ Lovable AI Gateway ↓ Google Gemini processes request ↓ AI response returned to browser ↓ Text not intentionally stored
Summary table
| What you do | Data leaves browser? | Destination | Stored by Textorum.io? |
|---|---|---|---|
| Type text for analysis | Not intentionally | — | — |
| Readability checker | Not intentionally | — | — |
| Keyword density checker | Not intentionally | — | — |
| SEO score | Not intentionally | — | — |
| AI Rewrite / Optimize | Yes (HTTPS) | Supabase → Google Gemini | Not intentionally |
| AI metadata generation | Yes (HTTPS) | Supabase → Google Gemini | Not intentionally |
| Visit any page | Standard server logs | Cloudflare / Lovable hosting | Temporarily (connection metadata only) |
Open questions & transparency
We believe privacy means being honest about what we do and do not control. AI requests processed through Google Gemini and the Lovable AI Gateway are governed by their respective terms and data retention policies. We do not control those retention policies and recommend reviewing them if this matters to you:
Last updated
- Last updated: May 2026
- Next scheduled infrastructure review: November 2026
Note: This page is updated whenever our infrastructure or data handling changes materially.
Contact
If you have technical questions about how Textorum.io processes data, contact us at jakob@kamender.com. For the full legal text, see our Privacy Policy.