AI 产品解读Provides a sidebar and web UI for interacting with locally running AI models (Ollama, Chrome AI/Gemini Nano, OpenAI-compatible APIs) directly from the browser
Users running local AI models lack a convenient, integrated browser interface to chat with their models while browsing. Page Assist bridges this gap by offering a sidebar and full web UI without requiring cloud services or accounts.
User installs extension → runs a local AI model server (e.g., Ollama) → opens the sidebar (Ctrl+Shift+Y) or web UI (Ctrl+Shift+L) → chats with the local model → can also summarize YouTube videos, chat with PDFs, use internet search, or download new models from ollama.com/huggingface.co directly via integrated buttons.
Sidebar chat interface for local AI modelsFull web UI accessible via options pageSupport for Ollama, Chrome AI (Gemini Nano), and OpenAI-compatible APIs (LLaMA.cpp, LM Studio, Llamafile, vLLM)Chat with PDF and documents (PDF, CSV, TXT, MD, DOCX)Internet search integrationYouTube video summarization via content script
- 目标用户
- Developers and power users running local AI models / Privacy-conscious users who prefer local inference over cloud APIs / Ollama users wanting a browser-integrated UI / Users who want to chat with documents and web pages using local models
- Source code is heavily minified/bundled (8.5MB PageAssistProvider chunk), making deep code inspection difficult
- No UI screenshots available to visually confirm feature gating or paid tiers
- The extension's own settings/options UI could not be rendered to check for hidden premium prompts
付费分析未识别到付费功能
No paid features, subscription tiers, or payment flows exist in this extension. The deterministic heuristic found zero paid signals. The extension is open-source (GitHub: https://github.com/n4ze3m/page-assist) and free. All 'pro' keyword hits are false positives from minified library code. No Stripe, billing, premium, upgrade, or payment-related code was found. The extension interfaces with user-hosted local AI models and user-provided API endpoints with no gating of features behind payment.
- 置信度
- 95
- 支付平台
- --
- 来源
- AI / 高
- 需要登录
- 否
- 登录理由
- The extension operates entirely locally with no account system. It connects to user-hosted local AI model servers (e.g., Ollama at 127.0.0.1:11434) and supports user-provided OpenAI-compatible API endpoints. No login, signup, or registration flow was found in any source file. The 'auth' keyword hits in background.js are all from bundled library code (Zod schema validation, JSON Schema), not authentication logic.
Работает. Насколько хорошо, зависит от модели и ваших настроек. Если попросить, выдает реальные результаты поиска в google со ссылками, и коротко пересказывает их. Ну, или читает за вас здоровенную страницу текста, переполненного саморекламой, и потом отвечает, есть там вообще что-то о том, зачем вы сюда пришли, или это очередная тупость.
Well polished, feature packed
I’ve been using this plugin for a while now, and I’m really impressed with how well it works. It’s fantastic at summarizing and handles a variety of tasks effortlessly. It’s also super fast, and the clean, user-friendly interface makes it a joy to use. That said, I think it could be even better with a couple of additions. For example, it would be awesome if you could attach an image to a system prompt, and have it display alongside it. Another cool feature would be the ability to chat with multiple models at the same time, which would add a lot of flexibility. Other than that, this is an amazing extension, and I highly recommend it to anyone!
Love this extension. Compared to AnythingLLM and OpenwebUI, both of them frequently time out when using with Ollama, but PageAssist gets it right, doesn't waste valuable time spent composing a well crafted prompt to have it lost because it "times out". You can save and export chats and settings, and manage a full array of LLM settings and models. The only feature I would like to see is having the ability to configure multiple remote servers/LLM's so I can chat with different models at the same time (while waiting for a response).
Works great, thanks!
ممتاز جدا لكن للأسف لا يدعم اللغة العربية
Great extension, but the sidebar functionality doesn't seem to work when using the Arc browser in MacOS. I left click somewhere on the page -> Open Copilot To Chat and nothing happens! :(
Wow, really nice extension, it feels fine to work with local llm and ollama in a useful and practical way. thanx
I like this application. I wish it could use two different URLs or two different ports in the OLLama URLs. That would allow me to utilize multiple PCS.
Congrats and thanks for bring us this extension. Fantastic