ColibotAI

Product Introduction

1. Definition: ColibotAI is a privacy-first, open AI engine Chrome extension designed for in-browser text processing. It falls under the category of browser productivity tools and client-side AI applications.
2. Core Value Proposition: It solves the core privacy and lock-in problem of AI browser extensions by giving users complete control over their data and AI provider. ColibotAI enables users to translate, summarize, and explain any web text using either Chrome's free on-device AI (Gemini Nano) or their own API key/model, ensuring no data is tracked, stored, or sent to a third-party backend.

Main Features

1. Hybrid AI Engine Architecture: This is the core technical innovation. Users can select their processing mode: (1) On-device Mode: Uses Chrome's built-in Gemini Nano model via the Built-in AI API. All processing occurs locally on the user's device; data never leaves the computer. This mode is free, private, and works offline for supported languages. (2) User's Key Mode: The extension sends the selected text directly from the browser to a user-configured endpoint (e.g., OpenAI, Anthropic Claude, Google Gemini, OpenRouter) or a local OpenAI-compatible endpoint (like Ollama, LM Studio). ColibotAI acts as a pure client; it has no backend server and never sees or stores the data or API keys, which are managed locally.
2. Core Text Processing Actions: The extension provides one-click actions on selected text: Translation across a wide language set, Smart Summarization to condense passages into key points, and Detailed Explanation for contextual understanding of complex content. A "Search with ChatGPT" option sends the selection to a new ChatGPT tab for deeper analysis.
3. Privacy and Local-First Data Management: All user data, including API keys and saved notes, is stored exclusively in the browser's local storage. The extension adheres to a strict no-analytics, no-cookies, no-account policy. Notes are saved locally, searchable within the extension's side panel, and can be exported or deleted by the user at any time. It is GDPR-compliant and Made in Switzerland.

Problems Solved

1. Pain Point: The fundamental privacy risk and vendor lock-in associated with most AI browser extensions, which typically require accounts, run analytics, and route all user data through their proprietary servers.
2. Target Audience: Privacy-conscious professionals, students, researchers, developers, and digital workers who regularly process web content but are concerned about data privacy, cost, and model flexibility. This includes users who want to leverage local AI models (Ollama, LM Studio) or utilize their existing API subscriptions (Claude, GPT, Gemini) without middlemen.
3. Use Cases: Real-time translation of foreign language articles during research; summarizing lengthy reports or documentation for quick comprehension; explaining technical passages or academic texts; converting web content into actionable local notes for personal knowledge management; and utilizing local LLMs for cost-effective, private text processing.

Unique Advantages

1. Differentiation: Unlike traditional AI extensions, ColibotAI does not function as a data conduit. It is a client-side tool with no central server. Competitors often require login, inject tracking, and store user queries; ColibotAI eliminates all such vectors. It also differentiates by supporting local/custom OpenAI-compatible endpoints, a feature rare in mainstream extensions.
2. Key Innovation: The Hybrid AI Engine with explicit, user-controlled routing of data. The architecture is designed around the principle of data minimization and local processing. The Token-saving reuse feature is a technical innovation that compares new text to previous queries and reuses saved answers when similarity exceeds a user-defined threshold, directly reducing API costs and latency. The Zero-token key validation provides instant, private feedback on API key validity without making a billable API call.

Frequently Asked Questions (FAQ)

1. Is ColibotAI really free, and how does it handle privacy with API keys? The extension is completely free to use. The on-device AI (Chrome's Gemini Nano) is also free. If you use your own API key, you pay your provider's usage rates directly. For privacy, your API key is stored locally in your browser and never leaves your device; it is sent directly to your chosen provider, with ColibotAI having no backend server to intercept or store it.
2. How does ColibotAI's privacy model differ from other AI browser extensions? Most extensions route your data through their servers for processing or logging. ColibotAI has no server or backend. In on-device mode, text stays on your computer. In "your key" mode, text is sent straight from your browser to the AI provider you select. No tracking, analytics, cookies, or accounts are used.
3. Can I use ColibotAI with a local LLM like Ollama? Yes. This is a core feature. ColibotAI can connect to any OpenAI-compatible API endpoint, including locally running models via Ollama, LM Studio, or llama.cpp. This enables completely offline, private, and free AI processing using your own hardware.
4. What does "Made in Switzerland" and "GDPR compliant" mean for this product? It signifies that ColibotAI is developed under Swiss jurisdiction, which has strong data protection laws. Being GDPR compliant means the extension's architecture is designed to respect user data rights by default—enforcing data minimization, providing transparency about processing (you see exactly where data goes), and giving users control over their data and consent.
5. How does the "Token-saving reuse" feature work to reduce costs? When you process text, ColibotAI checks it against a local cache of previously answered queries. If the new text is sufficiently similar to a saved query (based on a user-configurable similarity threshold), it returns the cached answer immediately without calling the AI API. This prevents spending tokens on near-duplicate requests, lowering costs and speeding up repeated workflows.