Opera Launches WebAI and HTML-in-Canvas, Bringing Local AI to the Browser

On October 27 2026, Opera announced that its flagship Opera One browser now includes an early-preview implementation of WebAI, a proposed web standard that lets websites and extensions run AI models directly on the user’s device. The same update also introduced HTML-in-Canvas, which embeds fully interactive web pages inside 2D and 3D scenes. Both features are opt-in and initially available on 64-bit Windows PCs with Intel or AMD processors and on Macs with Apple silicon. Opera GX and Opera Neon are slated to receive the same functionality soon.

Opera Launches WebAI, Bringing Local AI Models to Its Browsers 9to5MacMakeUseOfTechCrunch

Opera began experimenting with on-device large language models (LLMs) after an April 3 2024 announcement that added experimental support for roughly 150 local LLM variants from about 50 model families. Those early builds required 2 to 10 GB of storage per model and complemented Opera’s cloud-based AI service, Aria. The new WebAI rollout builds on that work, emphasizing open-weight models—such as Google’s Gemma and Apple’s on-device foundation model—that are free to download.

Opera Launches WebAI, Bringing Local AI Models to Its Browsers

• April 3 2024 – Opera announces experimental support for 150 local LLM variants.

Opera Launches WebAI, Bringing Local AI Models to Its Browsers

• October 27 2026 – Opera ships WebAI and HTML-in-Canvas as an early preview in Opera One, with plans to extend to Opera GX and Opera Neon.

Opera Launches WebAI, Bringing Local AI Models to Its Browsers

• Model support: Google’s Gemma and Apple’s on-device foundation model are available at launch; the implementation is model-agnostic.

Opera Launches WebAI, Bringing Local AI Models to Its Browsers

• Storage requirements: Individual local models typically need 2–10 GB of disk space.

Opera Launches WebAI, Bringing Local AI Models to Its Browsers

• Platform availability: 64-bit Windows PCs (Intel/AMD) and Apple-silicon Macs; not supported on Intel-based Macs, Linux, or Windows on Arm.

Opera Launches WebAI, Bringing Local AI Models to Its Browsers

• APIs: WebAI includes eight built-in APIs—Prompt, Decision, Semantic Embeddings, Summarizer, Writer, Rewriter, Translator, and Proofreader—plus a developer playground.

MakeUseOf

• HTML-in-Canvas: Enables live, interactive HTML/CSS elements to function within 2D/3D scenes, preserving full interactivity.

Opera Launches WebAI, Bringing Local AI Models to Its Browsers

• The feature is opt-in, users control which sites can access local models, and no AI subscription is required.

TechCrunch

• Opera’s developer blog notes that downloading a compatible local model may take several minutes and performance depends on hardware. The feature does not require a separate AI app, local server, or API key.

Opera Launches WebAI, Bringing Local AI Models to Its Browsers

• Privacy: Inference runs on the device, keeping prompts and generated content local.

Opera Launches WebAI, Bringing Local AI Models to Its Browsers

• Cost: Developers add AI capabilities without per-request fees, and users avoid subscription fees.

9to5Mac

• Innovation: Local AI combined with HTML-in-Canvas enables interactive 3D web experiences—such as in-game shops or immersive dashboards—using standard web tools.

Opera Launches WebAI, Bringing Local AI Models to Its Browsers

• Ecosystem: As a proposed standard backed by Google and Microsoft, WebAI could influence broader browser adoption of on-device AI.

9to5Mac

Opera has released example extensions—Project Lens, Got It?, and Questionable Intelligence—to illustrate WebAI uses. The company plans to add more open models and developer tools, and to expand support beyond the current Windows and Apple-silicon platforms.

Opera Launches WebAI, Bringing Local AI Models to Its Browsers