Right AI

AI sidebar with a browser agent and parallel multi-model chat. Ask once, compare answers from many models side by side.

RightAI — AI sidebar, translator, and browser agent RightAI turns Chrome into a productivity hub for AI work. Chat with any model side-by-side with the page you're reading, translate selected text in place, run a browser agent that can click, read, and write for you, and keep your favorite AI sites one shortcut away. Talk to AI without leaving the page - Right sidebar chat — A clean, resizable side panel for multi-turn AI conversations, with full page context (title, URL, main content) injected on demand and optional web fetch for up-to-date answers. - Multi-provider — Works with any OpenAI-compatible endpoint, plus first-class support for BigModel (Zhipu), local models, and custom bases. Pick the model you want per task; "endpoint models" appear in the picker when the server exposes them. - Quick Input — Press Ctrl+Q anywhere to send a prompt with the current page attached. Choose between a floating page overlay or a dedicated conversation window. - Attachments & images — Drag in files, paste screenshots, or right-click any image to OCR it inline. Omni — your browser agent Omni is an autonomous agent that drives a real browser to finish multi-step tasks. It runs locally in the extension on top of an OpenAI-compatible model (Click ✅ in settings for a free OpenCode Zen walkthrough). - Reading & writing — read_file , write_file , and diff_file for local file work, with an authorized directory workspace and per-import file selection. - Page interaction — screenshot , run_javascript (page-scoped, confirmation-gated), open_tab , switch_tab , and list_tabs to navigate without leaving the workflow. - Compound tasks — perform_task executes multi-step goals end-to-end, with streaming "thinking" and progress indicators you can follow live. - Safety first — Every privileged action asks for confirmation in the side panel; run_javascript is page-scoped and cannot touch Chrome APIs or local files. Translate, dim, focus - Selection translation — Select text on any page and get an inline bubble translation (also available in the right sidebar with page context). - Image OCR — Right-click any image for a floating OCR overlay with copy-to-clipboard. - Night Dim — Press Ctrl+Shift+L to soften bright pages for late-night reading. - Page telemetry — Optional crash and Jenkins console shortcuts for when you need them. Built around your workflow - Customizable toolbar — Pin any AI site (ChatGPT, Claude, Gemini, Grok, DeepSeek, Doubao, Kimi, and 30+ more) or your own custom links. Drag to reorder, import/export your layout. - Flexible layouts — Top dock or right-side panel; optional split view for working with two panes at once. - 47 languages — Fully localized UI including English, Chinese, Japanese, Korean, Spanish, French, German, Arabic, Hindi, and more. - Liquid-glass UI — Smooth, modern, theme-aware (light/dark) interface with persistent history, file audit logs, and graceful confirmations. Why RightAI - Your keys, your data — API keys and settings stay on your device. No telemetry, no third-party middleman. Configuration is stored locally via chrome.storage and localStorage with import/export. - No bundler, no bloat — Native ES modules, MV3-compliant, lean permission set (declared in manifest.json ). - Transparent — Every powerful action shows you exactly what is about to happen and waits for your go-ahead. Permissions - Read & modify page content (for translation, OCR, agent actions, side-panel context). - Storage (for settings, history, and the authorized local workspace). - Tabs (for agent tab management, Quick Input scoping, and split view). - Notifications and <all_urls> scripting are minimized to what each feature actually needs. Open the side panel with Ctrl+R and start exploring.