Privacy Policy

RenameClick is a local-first application. Your files are processed entirely on your device, and we never copy, analyze, or store their contents on our servers. This policy explains the limited information we handle to run the product.

1. Information Processed Locally

Everything you import, rename, or analyze within the app stays on your machine. Cloud services are not used for file processing unless you explicitly connect to one.

2. API Keys and Third-Party AI Services

RenameClick allows you to optionally connect your own API keys for OpenAI (ChatGPT) and Google (Gemini) to use cloud-based AI features. Here is how we handle this data:

• Local storage only: Your API keys are stored exclusively on your device using Electron's secure storage, which leverages OS-level encryption (Keychain on macOS, Credential Vault on Windows, libsecret on Linux).
• Never transmitted to us: Your API keys are never sent to RenameClick servers. We have no access to your keys or the ability to retrieve them.
• Direct communication: When you use cloud AI features, your device communicates directly with OpenAI or Google servers. RenameClick acts only as a local interface — we do not proxy, intercept, or store any data exchanged with these providers.
• Your data, your control: Any document content sent to third-party AI providers is transmitted directly from your device. We have no visibility into or access to your API requests or responses.

3. Information We Collect

We only capture essential account data:

• Billing data (Pro users): Polar.sh securely processes your payment method and retains the minimum detail required for receipts and compliance. We only receive subscription status and invoice metadata.
• License records: We keep your license key, plan status, and contact email so we can authenticate the app and provide support.
• Crash reporting (opt-out): We use Sentry to collect anonymous crash reports, enabled by default. These reports contain only crash stack traces and app version — no personal data or file contents are ever collected. You can disable crash reporting at any time in the app settings.

4. How We Use Information

• Authenticate devices and deliver Pro features you're entitled to.
• Process subscriptions, invoices, and basic communications.
• Respond to support requests and improve the product based on anonymized trends.

5. Third-Party Services

We partner with Clerk.com for authentication and account session management. They process identifiers such as your email, device information, and sign-in events so we can securely log you in. Polar.sh manages subscription billing, invoices, and payment confirmations. Sentry.io receives anonymous crash reports (when enabled) containing only technical stack traces and app version — no personal information or file data is transmitted.

If you choose to use cloud AI features with your own API keys, your device connects directly to OpenAI and/or Google. These connections are initiated by you and governed by your own accounts with these providers. We do not intermediate, log, or have any access to these communications. Your use of OpenAI and Google services is subject to their respective privacy policies, which we encourage you to review.

Each provider stores only what is necessary for their role and operates under their own privacy policies. We do not permit these vendors to use your information for unrelated purposes.

6. Storage and Security

Account metadata is stored in encrypted systems with access limited to essential personnel. We retain it only while you have an active relationship with us or as required by law. Backups follow the same controls.

7. Your Choices

• Manage or cancel subscriptions via the billing portal at any time.
• Disable crash reporting inside the app settings if you prefer not to share anonymous diagnostics.
• Contact us at support@pithly.app for privacy questions or to request data deletion.

8. Local AI Model Transparency

RenameClick includes an optional offline AI model that runs entirely on your device. No files or content are uploaded or transmitted when using local inference.

This architecture is designed for privacy-sensitive workflows, including scenarios where users handle regulated data (e.g. healthcare or legal documents), when used in local-only mode. We use unmodified, openly licensed models and provide full transparency for verification.

Current Model: Qwen3-VL-4B-Instruct (Q4_K_M Quantization)

This vision-language model is published by Unsloth under the Apache 2.0 license. We distribute unmodified copies of the original model files.