How We Removed Paywalls from Our Slate for Rise Chrome Extension

Slate for Rise is our free Chrome extension for customizing Articulate Rise courses. It lets you inject course-wide CSS and JavaScript, preview changes live, and apply your customizations automatically when you export to any format (SCORM, xAPI, Web, you name it). No unzipping files, no manual editing.

With the latest update, we've added a free AI code assistant. No subscription, no API keys, no account. It's completely optional—you choose whether to enable it, and all the other extension features work without it.

We made this possible by running a small language model directly in your browser. It's a good example of when local AI makes sense.

Small models for small tasks

Not every AI task needs a massive model with billions of parameters. Writing a CSS animation, explaining a JavaScript snippet, or optimizing a selector—these are focused, well-defined problems. A smaller model can handle them well.

We use Meta's Llama 3.2 1B Instruct, a compact model designed for exactly this kind of work. It runs entirely in your browser using WebGPU, a relatively new API that lets web applications tap into your computer's GPU. The first time you use the AI features, you'll download the model files (around 700MB) to your browser's cache. After that, everything runs locally.

The trade-off with local models is speed. Response time depends on your machine—a modern laptop with a decent GPU handles it smoothly, while older hardware might see a short delay. But for the focused tasks in our code editor, it works well.

When you need more horsepower

Small models have limits. They're great for targeted code tasks, but they can't handle complex multi-step reasoning, large context windows, or intricate data structures.

That's why our other product, Slate (our full eLearning authoring tool), uses Gemini from Google for its AI features. Course authoring involves generating lesson content, understanding complex document structures, handling translations, and working with large amounts of context. Those tasks genuinely need a more powerful model running in the cloud.

The point isn't that local AI is always better—it's about matching the tool to the task. For CSS and JavaScript assistance in a Chrome extension, a small local model is the right fit. Especially when your goal is eliminating costs for users. That was our challenge with Slate for Rise: could we make the entire thing free? Turns out, yes.

What you can do with it

The AI assistant helps with the CSS and JavaScript you're writing for your Rise courses:

• Generate code from plain language descriptions ("make the continue button float up and down")
• Explain code so you understand what a snippet does before using it
• Optimize code for better performance or cleaner syntax

It's built right into the code editor. Select some code, hit a keyboard shortcut, and you're working with the AI. No context switching, no copy-pasting into a separate tool.

Privacy as a bonus

Running the model locally means your code never leaves your machine. There's no telemetry, no analytics on what you're writing, no data collection. For enterprise teams working on internal training content, that's a real benefit.

Get started

If you already have Slate for Rise installed, update to the latest version. The AI features are in the code editor toolbar. First use will prompt you to download the model.

New to Slate for Rise? Install it free from the Chrome Web Store. You get course-wide CSS and JavaScript injection, live preview, bulk processing, and free AI code assistance.

The model download is optional. All the other features work without it.

Have feedback on this update? We'd love to hear from you.