Chrome Skills vs Gemma Gem

Google launched Skills for Chrome on April 14, 2026, bringing reusable AI workflows directly into the browser sidebar. The core idea is deceptively simple: any Gemini prompt you find useful can be saved as a "Skill" and triggered later with a forward slash (/) command — no copy-pasting, no re-explaining context. You can also run a Skill across multiple tabs simultaneously, or remix community Skills from Google's growing library of pre-built workflows. The Skills library covers categories like productivity, shopping, recipes, and budgeting. Power users can build multi-step workflows — summarize, translate, then draft a reply — and trigger the whole chain with a single command. Privacy-sensitive actions (adding calendar events, sending emails) require explicit confirmation. The rollout began on macOS, Windows, and ChromeOS for English-US users signed into Gemini. This matters because it's the first time a major browser has made AI-native workflows a first-class citizen, not a plugin or extension. It's also a quiet shot across Perplexity, Copilot, and any browser extension trying to bolt AI onto the web. If you're already in the Google ecosystem, this starts to make the browser feel like an operating system.

Gemma Gem is an open-source Chrome extension that runs Google's Gemma 4 language model entirely in your browser using WebGPU — no API keys, no server, no data leaving your device. Install the extension, wait for the one-time model download (500MB for the efficient 2B variant, 1.5GB for the larger 4B), and you have a fully private AI assistant that can read web pages, fill forms, take screenshots, and execute JavaScript. The extension uses Hugging Face Transformers.js with ONNX-quantized versions of Gemma 4's E2B and E4B variants, making the model small enough to run in a browser tab without throttling GPU memory. Gemma 4's strong efficiency profile — particularly its per-layer attention architecture — makes it a natural fit for WebGPU's memory constraints compared to older models at similar parameter counts. What makes Gemma Gem interesting beyond the cool factor: it's a glimpse at what fully private, zero-latency browser-native AI looks like. There's no round-trip to a server, no API billing, no rate limits. On a mid-range MacBook M3 or gaming GPU, inference is fast enough to be genuinely useful. The trade-off is capability — Gemma 4 E2B is a 2B parameter model, not Claude or GPT-5, but for summarization, form-filling, and basic Q&A it holds its own.