Transparency
Last Updated: August 9, 2026
Overview
This page discloses information about data processing, technologies used, and pricing at Land. For detailed privacy information, please refer to our Privacy Policy. For pricing details, please refer to our Notation Based on Specified Commercial Transaction Act.
Data Processing
File Upload Processing Flow
- Storage: Files are stored in Cloudflare R2 (encrypted at rest)
- Metadata Generation: AI analyzes the file to generate tags, categories, and a suggested name
- Text Extraction: Text content is extracted from documents and images (OCR for images)
- Embedding: Content is converted to vector embeddings for semantic search
- Placement: The file is positioned on the canvas near semantically similar files
Data Retention
- Files are stored until the user deletes them
- Deleted files are permanently removed within 30 days
- Account deletion removes all associated data
For full data retention details, please refer to our Privacy Policy.
AI Processing
Technologies Used
Land performs AI processing across the following categories. The specific models used vary depending on your plan and the type of processing, and are updated regularly to improve service quality.
| Task | Provider | Notes |
|---|---|---|
| Text Generation & Summarization | Cloudflare Workers AI | Multiple models used depending on task |
| Vision & OCR | Cloudflare Workers AI | Vision-capable models |
| Transcription | Cloudflare Workers AI | Audio file transcription |
| Advanced Text Processing | Google AI (via AI Gateway) | Available on Pro plan and above |
| Embeddings | Cloudflare Workers AI | For semantic search |
The latest list of available models can be found in the settings within the service.
External AI Routing by Plan
Which providers process your content depends on your plan and, when set, your Bring Your Own Key (BYOK) preferences.
| Plan | On-platform (Cloudflare Workers AI) | External via Cloudflare AI Gateway |
|---|---|---|
| Free | Metadata / Summary / OCR / Embeddings | None by default |
| Pro | Metadata / Summary / OCR / Embeddings / Transcription (Whisper) | Google Gemini (text, OCR); OpenAI GPT-4o Transcribe (transcription fallback) |
| Team | Same as Pro | Same as Pro |
Transcription is not available on the Free plan by default. BYOK, when configured for a model, routes that specific task to the chosen provider (OpenAI, Anthropic, or Google) with your own key; that provider's terms of use apply.
Data Usage
- User data is never used to train AI models. Land's own policy is reinforced by each provider's contract:
- Cloudflare Workers AI: Data usage — customer content is not used for training
- Cloudflare AI Gateway (proxy for external calls): Data control — request logging is under the account owner's control
- Anthropic Commercial API: Commercial Terms of Service — Anthropic does not use Customer Content submitted to its commercial API to train its models
- Google Gemini paid API (via AI Gateway): Terms of Service — paid-tier requests are excluded from training
- OpenAI (via AI Gateway or BYOK): API Data Usage Policy — API traffic is not used to train models
- AI processing runs on Cloudflare's infrastructure (Workers AI directly; AI Gateway proxies external providers)
- BYOK (Bring Your Own Key) option available for OpenAI, Anthropic, and Google models. When BYOK is used, the account owner's provider-side terms apply and Land does not add its own retention.
AI-generated Content Disclosure
Where Land displays AI-generated fields (tags, summary, suggested filename), the UI shows a small "AI" indicator so you can distinguish generated content from your own input.
Token Pricing
Tokens are the unit of AI processing. Each operation consumes tokens based on the model used.
| Plan | Monthly Tokens | Approx. File Operations |
|---|---|---|
| Free | 600,000 | ~100 files |
| Pro | 8,000,000 | ~1,100 files |
| Team | 16,000,000 | ~2,200 files |
Pricing is based on a unified rate of $0.20 per 1 million tokens, with per-model cost equivalents. For detailed plan pricing, please refer to our Notation Based on Specified Commercial Transaction Act.
Infrastructure
| Component | Provider | Purpose |
|---|---|---|
| API Server | Cloudflare Workers | Backend processing |
| Database | Cloudflare D1 (SQLite) | Metadata storage |
| File Storage | Cloudflare R2 | File storage (S3-compatible) |
| Vector Search | Cloudflare Vectorize | Semantic search |
| AI Processing | Cloudflare Workers AI | On-platform AI |
| AI Gateway | Cloudflare AI Gateway | External AI routing |
| CDN | Cloudflare | Content delivery |
Security
We implement data encryption in transit (TLS) and at rest, CSRF protection, rate limiting, and secure session management. For details on our security measures, please refer to our Privacy Policy.
Third-party Services
| Service | Purpose | Data Shared |
|---|---|---|
| Cloudflare | Infrastructure, CDN, AI processing | User Content, usage data |
| Google (OAuth, Gemini) | Authentication, AI processing | Authentication data, data for processing |
| Stripe | Payment processing | Payment information (not held by us) |
| Google Analytics 4 | Usage analysis | Anonymized usage data |
| Sentry | Error monitoring | Error information, device information |
For details on data sharing with third-party services, please refer to our Privacy Policy.
Experiment Assignment ID (A/B testing)
Land runs product-improvement experiments (A/B tests) to evaluate new features and UI changes. To keep your assignment stable across visits — so the experience does not switch between variants every time you reload — we store a long-lived anonymous identifier named land_aid in a cookie or in localStorage as a fallback. land_aid contains no personally identifying information and is never shared with third parties. See the Cookie Policy for details.
Session Identifier for Usage Measurement
To know which actions belong to the same single visit, we store a randomly generated session identifier in your browser. It contains no personally identifying information, is lost when you close the tab, is not carried over to your next visit, and is never shared with third parties.
We use it so that activity before signing in and activity after signing in can be recognised as one visit. Without it we cannot count something as basic as how many of the people who tried the demo went on to sign up, which would leave us guessing about where the experience breaks down.
If you opt out of analytics in your settings, we do not retain this identifier. It is distinct from land_aid and never persists across visits. See the Cookie Policy for details.
Contact
Questions about data processing? Contact us at the Contact page.