Five AI tools that do the heavy lifting inside your browser tab. No API key required, no waitlist, no account, no data sent to a remote server. The models download once and run on your device using WebGPU or WebAssembly.
| Tool | Key Benefit | Time to Complete | Privacy Level |
|---|---|---|---|
| AI Text Summarizer | Condense articles and reports | < 10 seconds | 100% Local |
| Text Humanizer | Make AI-written text sound natural | < 10 seconds | No Data Sent |
| Grammar Checker | Fix grammar and style | < 5 seconds | Zero Server Uploads |
| Resume Builder | Generate a formatted resume | < 2 minutes | Files Stay on Your Machine |
| AI Prompt Builder | Craft effective prompts for any AI | Instant | 100% Local |
1. AI Text Summarizer — Cut Through Long Content Fast
AI Text Summarizer condenses articles, reports, emails, and research papers into a short, medium, or detailed summary. The model runs via Transformers.js on WebGPU — processing happens on your hardware, not on a remote API.
Who it’s for: Researchers processing large volumes of literature, business professionals sifting inboxes, students reviewing lecture notes before exams.
The technical detail: Uses a client-side NLP pipeline with env.allowRemoteModels = false enforced — the model cannot make outbound calls once loaded. Your text is processed entirely in browser memory.
2. Text Humanizer — Make AI Output Sound Like You
Text Humanizer rewrites AI-generated text to sound more natural, varied, and conversational — removing the telltale patterns that make AI writing detectable. Paste any text, and the browser-side model rewrites it.
Why this matters: AI detectors are increasingly used by publishers, educators, and editors. Humanizing text locally means your content is never passed through a third-party rewriting API that could log your input.
Performance tip: First run takes 10–20 seconds to load the model into memory. Subsequent uses in the same session are near-instant.
3. Grammar Checker — Instant Proofreading in the Browser
Grammar Checker identifies and corrects grammar, punctuation, and style issues using a lightweight on-device model. Unlike cloud-based grammar tools, your text is never transmitted — an important difference when proofreading confidential documents, contracts, or personal correspondence.
Who it’s for: Writers, students, non-native English speakers, professionals editing client-facing documents.
4. Resume Builder — Generate a Polished CV Locally
Resume Builder uses a Llama 3.2 model running in the browser to help you write and format a professional resume from your raw inputs. Built with Meta Llama — see the “Built with Llama” attribution in the tool.
Privacy note: Your work history, contact details, and career information are among the most sensitive personal data you have. The Resume Builder processes everything locally — nothing is uploaded to a server or shared with a third party.
Output formats: The finished resume can be downloaded as a formatted document directly from the browser.
5. AI Prompt Builder — Get Better Results from Any AI
Prompt Builder helps you craft effective prompts for ChatGPT, Claude, Gemini, and other AI systems. Describe your goal, choose a tone and format, and the tool outputs a structured prompt ready to paste into any AI interface.
Who it’s for: Anyone who regularly uses AI tools and wants to move beyond basic one-liner prompts. Better prompts produce better outputs — this tool runs locally and helps you get there faster.
What to Do Next
Combine these AI tools with the rest of the ALLYX ONE toolkit for a complete workflow:
- Working with documents? → PDF Merger and PDF to Word handle document prep before you summarise or edit
- Need images processed? → Background Remover runs a separate on-device AI model for image segmentation
- Writing a YouTube script? → YouTube Video Summarizer extracts the key points from any transcript so you can rework the content
Frequently Asked Questions
Do I need an API key to use these AI tools? No. All five tools use on-device AI models (via Transformers.js + WebGPU) that run locally in your browser. No external API, no API key, no account.
How is this different from using ChatGPT or Claude? Cloud-based AI services process your text on their servers and may use inputs to improve their models (depending on settings). These tools run locally — your text is processed in your browser’s memory and never leaves your device.
Is the AI quality comparable to cloud services? On-device models are smaller than large cloud models and have different capabilities. They perform well for summarisation, grammar, and rewriting tasks. For very complex reasoning or long-form generation, a large cloud model may produce better results.
What happens the first time I open a tool? The AI model (typically 40–150 MB depending on the tool) downloads to your browser’s local storage. A progress bar shows download status. Subsequent visits load the model from local storage instantly.
Do these tools work on mobile? The tools load on mobile browsers. WebGPU acceleration for faster inference is available on Chrome for Android. iOS Safari has partial WebGPU support; the tools fall back to WebAssembly (slower but functional).