LevelWith — plain-English explainer

**Let me level with you.** The web is full of writing pitched at someone else. Medical sites assume you're a clinician. Financial pages assume you already know what a convexity adjustment is. Developer docs assume you've been in the ecosystem for years. LevelWith reads whatever you're looking at and explains it the way a smart, patient friend would — in plain English, at the depth you ask for, and tailored to who you actually are. **What makes LevelWith different** LevelWith has a lightweight "About Me" profile — your work, your hobbies, how you prefer to learn — and it uses that profile to shape every explanation. If you're a dentist reading a paper about AI, the analogies come from the operatory. If you're a woodworker reading about machine learning, the examples come from the shop. If you're a parent trying to understand a mortgage document, the scenarios are grounded in your life, not a generic "a business might…" template. You can still use LevelWith without filling in a profile — it just gets noticeably sharper when you do. **Four ways to ask** - **Explain this page** — one click on any article, docs page, or blog post - **Explain this selection** — highlight any paragraph, right-click, and get a plain-English read - **Explain this screenshot** — drop in an image of a chart, slide, or document (vision-capable models) - **Just ask a question** — type anything directly, like "what is a SAFE note, and when would I actually care?" **Pick your depth** Every explanation comes with four depth settings: **ELI5** (explain it to a curious kid), **Adult** (smart generalist, no jargon), **Pro** (working professional in the field), and **Expert** (you already know the basics and want the nuance). Switch depths on any result and LevelWith re-explains at the new level in place. **What you get back** Every explanation is structured — not a wall of text. You'll see a TL;DR, what the thing actually is, an analogy drawn from your profile, concrete ways you might use it, a jargon decoder for any terms that tripped you up, and (when the source warrants it) a "red flags" section calling out claims to be skeptical of. Plus suggested follow-up questions and a chat box for anything else. **Your glossary grows with you** Every time LevelWith decodes a jargon term for you, it adds it to a personal glossary you can search later. Your own custom dictionary of everything you've learned, searchable from the popup. **Bring your own API key — you stay in control** LevelWith works with Anthropic (Claude), OpenAI (GPT), and Ollama (local or Ollama Cloud). You supply the API key; LevelWith talks directly to the model provider from your browser. We don't operate a server. We don't proxy your prompts. We don't see what you're reading. Your data goes from your machine to your chosen model provider and nowhere else. **Privacy-first by design** - No LevelWith backend, no telemetry by default, no analytics SDKs - Your profile and glossary live in your browser's local storage - All network traffic goes directly to the LLM provider you chose - Full source code available for review **Who it's for** - Learners reading above their pay grade who want context without dumbing things down - Professionals crossing into a new domain (doctors reading finance, engineers reading legal, PMs reading research papers) - Anyone who's tired of Googling acronyms mid-article - Researchers and analysts who want a second pair of eyes on dense sources **Getting started** 1. Install LevelWith 2. Open the popup and paste an API key from Anthropic, OpenAI, or your Ollama instance 3. Spend 30 seconds filling in "About Me" — work, hobbies, how you prefer analogies 4. Click the extension on any page and hit Explain Learn more: **https://levelwith.io** Questions, feedback, or bug reports: **support@levelwith.io**