crossmem
Harness your browsing into a personal wiki. One-click AI summaries with PDFs, screenshots, and markdown notes.
As of June 2026, crossmem has 3 users in the Productivity category.
Usersup 200.0 percent+200.0%
3
3
Ratingno change0%
—
— reviews
Reviewsno change0%
—
Version
1.1.0
Manifest V3
90-day change · In the last 90 days this extension 2 version updates, changed permissions.
History
6 snapshotsTracking since Apr 11, 2026.
View as table
| Date | Users | Rating | Reviews | Version |
|---|---|---|---|---|
| Apr 11, 2026 | 1 | — | — | 0.9.0 |
| Apr 20, 2026 | 3 | — | — | 1.0.0 |
| Apr 27, 2026 | 3 | — | — | 1.1.0 |
| May 10, 2026 | 2 | — | — | 1.1.0 |
| May 22, 2026 | 3 | — | — | 1.1.0 |
| Jun 10, 2026 | 2 | — | — | 1.1.0 |
| Now | 3 | — | — | 1.1.0 |
Changelog
- Apr 20, 2026description
Snapshot & summarize any web page with AI. Your personal knowledge base — one memory, every AI. Detailed description: crossmem turns any web page into a memory. One click to capture, AI summarizes, and a memory ball is created — auto-categorized by topic, auto-titled, with the source favicon. 0% 🧠 HOW IT WORKS 1. Open any web page 2. Click "Snapshot & Summary" in the side panel 3. AI reads and summarizes the page into bullet points 4. A colored memory ball appears — purple for research, yellow for ideas, blue for tech, green for resources, red for action items 📂 STORAGE • Local: saves screenshots and markdown to ~/crossmem/{raw, wiki} • Google Drive: saves to your own Drive (optional) • arxiv papers: PDF downloaded automatically 🤖 TWO AI MODES 0% • Bridge Mode — uses your local Claude Code subscription (free, no API cost). Run: npx crossmem-bridge • Standalone Mode — use your own API key (Claude, OpenAI, or OpenRouter). No setup needed. 🔗 WORKS WITH • Obsidian — open wiki summaries directly in your vault • Any website — logged-in pages work too (uses your real browser session) • CLI agents — control the browser from your terminal via WebSocket bridge 📸 FEATURES • Memory balls with auto-color categorization • Markdown rendering in side panel • Source favicon on every memory • Open Raw (screenshot/PDF) and Wiki (markdown) separately • Settings: choose AI mode, API provider, cache limit • arxiv auto-detect and PDF download 🔒 PRIVACY • Your data stays local or in your own Google Drive • API keys stored locally, never sent to our servers • No data collection, no tracking crossmem.dev — one memory, every aicrossmem turns any web page into a structured memory — automatically. Click Snapshot on an arxiv paper, a Wikipedia article, or a blog post. crossmem reads the page, generates an AI summary, and files it as a memory ball in your side panel. The raw source (PDF or screenshot) and a wiki-style markdown note are saved to your local filesystem or Google Drive. HOW IT WORKS 1. Open any web page 2. Click Snapshot in the side panel (or right-click → Remember This Page) 3. AI summarizes the page and categorizes it (research, tech, resource, idea, action) 4. A memory ball appears with title, summary, source link, and category 5. Raw file (PDF/screenshot) + wiki markdown saved automatically WHAT YOU GET - Memory balls: color-coded by category, searchable, expandable - Raw snapshots: arxiv PDFs downloaded, other pages screenshotted - Wiki notes: structured markdown with frontmatter, ready for Obsidian or any editor - Side panel UI: always accessible, never leaves your browser - Right-click menu: "Remember This Page" or "Remember Selection" - Bridge API: local WebSocket API for scripts and AI agents to drive crossmem programmatically STORAGE OPTIONS - Local: files saved to ~/crossmem/ (configurable) via the companion bridge - Google Drive: summary files in your personal crossmem/ folder - Both: use local for raw files, Drive for cloud backup YOUR DATA, YOUR CONTROL - No backend servers, no analytics, no telemetry - AI calls go directly from your browser to your own API key - Everything stored locally or in your own Google Drive - Open source: inspect every line of code Built for researchers, developers, and anyone who reads more than they can remember. --- crossmem 把任何網頁自動變成結構化的記憶。 在 arxiv 論文、維基百科或任何 blog 上按一下 Snapshot,crossmem 讀取頁面、用 AI 產生摘要、在側邊欄生成一顆 memory ball。原始檔案(PDF 或截圖)和 wiki 風格的 markdown 筆記會自動存到你的本機或 Google Drive。 運作方式 1. 打開任何網頁 2. 在側邊欄按 Snapshot(或右鍵選 Remember This Page) 3. AI 摘要頁面並自動分類(research、tech、resource、idea、action) 4. 記憶球出現,帶有標題、摘要、來源連結、分類 5. 原始檔(PDF/截圖)+ wiki markdown 自動存檔 你會得到什麼 - 記憶球:依分類上色、可搜尋、可展開 - 原始快照:arxiv 論文下載 PDF、其他網頁截圖 - Wiki 筆記:帶 frontmatter 的結構化 markdown,可直接餵 Obsidian 或任何編輯器 - 側邊欄 UI:隨時可用,永遠不離開你現在的頁面 - 右鍵選單:Remember This Page 或 Remember Selection - Bridge API:本地 WebSocket,讓任何 script 或 AI agent 程式化操作 crossmem 儲存選項 - 本機:透過 bridge 存到 ~/crossmem/(可設定路徑) - Google Drive:摘要檔存在你自己的 crossmem/ 資料夾 - 混合:原始檔走本機、摘要走 Drive 作為雲端備份 你的資料、你的掌控 - 沒有後端伺服器、沒有數據分析、沒有 telemetry - AI 呼叫直接從瀏覽器走你自己的 API key - 所有資料都在本機或你自己的 Google Drive - 開源:每一行程式碼都可以檢視 為研究者、開發者、以及所有讀得比記得多的人打造。
- Apr 20, 2026short_description
Snapshot & summarize any web page with AI. One memory, every AI.
Harness your browsing into a personal wiki. One-click AI summaries with PDFs, screenshots, and markdown notes.
- Apr 20, 2026permissions
identity, storage, activeTab, scripting, sidePanel, alarms, tabs
identity, storage, activeTab, scripting, sidePanel, alarms, tabs, contextMenus
- Apr 11, 2026description
Save your AI conversations to Google Drive. Export from claude.ai and access them anywhere — your data, your cloud, your control.
Snapshot & summarize any web page with AI. Your personal knowledge base — one memory, every AI. Detailed description: crossmem turns any web page into a memory. One click to capture, AI summarizes, and a memory ball is created — auto-categorized by topic, auto-titled, with the source favicon. 0% 🧠 HOW IT WORKS 1. Open any web page 2. Click "Snapshot & Summary" in the side panel 3. AI reads and summarizes the page into bullet points 4. A colored memory ball appears — purple for research, yellow for ideas, blue for tech, green for resources, red for action items 📂 STORAGE • Local: saves screenshots and markdown to ~/crossmem/{raw, wiki} • Google Drive: saves to your own Drive (optional) • arxiv papers: PDF downloaded automatically 🤖 TWO AI MODES 0% • Bridge Mode — uses your local Claude Code subscription (free, no API cost). Run: npx crossmem-bridge • Standalone Mode — use your own API key (Claude, OpenAI, or OpenRouter). No setup needed. 🔗 WORKS WITH • Obsidian — open wiki summaries directly in your vault • Any website — logged-in pages work too (uses your real browser session) • CLI agents — control the browser from your terminal via WebSocket bridge 📸 FEATURES • Memory balls with auto-color categorization • Markdown rendering in side panel • Source favicon on every memory • Open Raw (screenshot/PDF) and Wiki (markdown) separately • Settings: choose AI mode, API provider, cache limit • arxiv auto-detect and PDF download 🔒 PRIVACY • Your data stays local or in your own Google Drive • API keys stored locally, never sent to our servers • No data collection, no tracking crossmem.dev — one memory, every ai - Apr 11, 2026short_description
Save AI conversations and web pages from Claude, ChatGPT, Gemini to your Google Drive
Snapshot & summarize any web page with AI. One memory, every AI.
- Apr 11, 2026host_permissions
(empty)
<all_urls>
- Apr 11, 2026permissions
identity, storage, activeTab, scripting, sidePanel
identity, storage, activeTab, scripting, sidePanel, alarms, tabs
Permissions & access
- Permissions
- identitystorageactiveTabscriptingsidePanelalarmstabscontextMenus
- Host access
- <all_urls>
Screenshots
About
crossmem turns any web page into a structured memory — automatically. Click Snapshot on an arxiv paper, a Wikipedia article, or a blog post. crossmem reads the page, generates an AI summary, and files it as a memory ball in your side panel. The raw source (PDF or screenshot) and a wiki-style markdown note are saved to your local filesystem or Google Drive. HOW IT WORKS 1. Open any web page 2. Click Snapshot in the side panel (or right-click → Remember This Page) 3. AI summarizes the page and categorizes it (research, tech, resource, idea, action) 4. A memory ball appears with title, summary, source link, and category 5. Raw file (PDF/screenshot) + wiki markdown saved automatically WHAT YOU GET - Memory balls: color-coded by category, searchable, expandable - Raw snapshots: arxiv PDFs downloaded, other pages screenshotted - Wiki notes: structured markdown with frontmatter, ready for Obsidian or any editor - Side panel UI: always accessible, never leaves your browser - Right-click menu: "Remember This Page" or "Remember Selection" - Bridge API: local WebSocket API for scripts and AI agents to drive crossmem programmatically STORAGE OPTIONS - Local: files saved to ~/crossmem/ (configurable) via the companion bridge - Google Drive: summary files in your personal crossmem/ folder - Both: use local for raw files, Drive for cloud backup YOUR DATA, YOUR CONTROL - No backend servers, no analytics, no telemetry - AI calls go directly from your browser to your own API key - Everything stored locally or in your own Google Drive - Open source: inspect every line of code Built for researchers, developers, and anyone who reads more than they can remember. --- crossmem 把任何網頁自動變成結構化的記憶。 在 arxiv 論文、維基百科或任何 blog 上按一下 Snapshot,crossmem 讀取頁面、用 AI 產生摘要、在側邊欄生成一顆 memory ball。原始檔案(PDF 或截圖)和 wiki 風格的 markdown 筆記會自動存到你的本機或 Google Drive。 運作方式 1. 打開任何網頁 2. 在側邊欄按 Snapshot(或右鍵選 Remember This Page) 3. AI 摘要頁面並自動分類(research、tech、resource、idea、action) 4. 記憶球出現,帶有標題、摘要、來源連結、分類 5. 原始檔(PDF/截圖)+ wiki markdown 自動存檔 你會得到什麼 - 記憶球:依分類上色、可搜尋、可展開 - 原始快照:arxiv 論文下載 PDF、其他網頁截圖 - Wiki 筆記:帶 frontmatter 的結構化 markdown,可直接餵 Obsidian 或任何編輯器 - 側邊欄 UI:隨時可用,永遠不離開你現在的頁面 - 右鍵選單:Remember This Page 或 Remember Selection - Bridge API:本地 WebSocket,讓任何 script 或 AI agent 程式化操作 crossmem 儲存選項 - 本機:透過 bridge 存到 ~/crossmem/(可設定路徑) - Google Drive:摘要檔存在你自己的 crossmem/ 資料夾 - 混合:原始檔走本機、摘要走 Drive 作為雲端備份 你的資料、你的掌控 - 沒有後端伺服器、沒有數據分析、沒有 telemetry - AI 呼叫直接從瀏覽器走你自己的 API key - 所有資料都在本機或你自己的 Google Drive - 開源:每一行程式碼都可以檢視 為研究者、開發者、以及所有讀得比記得多的人打造。
Technical
- Version
- 1.1.0
- Manifest
- V3
- Size
- 82.84KiB
- Min Chrome
- 88
- Languages
- 1
- Featured
- No
Metadata
- ID
- kmpfhoimimgfdglaglpjegjiahkfolpa
- Developer ID
- ue846070801019f1b32f10937c4a80905
- Developer Email
- [email protected]
- Created
- Feb 23, 2026
- Last Updated (Store)
- Apr 14, 2026
- Last Scraped
- Jun 10, 2026
- Website
- crossmem.dev
- Support URL
- —
- Privacy Policy
- https://yiidtw.github.io/crossmem-privacy/
Similar extensions
Alternatives to crossmem, ranked by description similarity.
Claude Chat Downloader
Download chat history, including all files. Save in multiple formats including JSON, HTML, and Markdown.
414
★ 5.0
Docify – Export Your AI Chats
The easiest way to save, export, and format your AI chat conversations.
82
★ 4.5
ConvoSnap: ChatGPT & Gemini to Google Docs | Export AI Chats to PDF
Export ChatGPT, Gemini & Claude chats to Google Docs, PDF, or Markdown. Save and organize AI conversations into folders.
1.0K
★ 3.8
Relai
Share AI conversations locally. No servers, no tracking, 100% private.
132
★ 4.0
AI Chat Exporter
Export and save conversations from ChatGPT, Claude, Gemini, and Grok
25
★ 5.0
Chat2Doc - Export ChatGPT, Claude & Gemini to PDF/Word
One-click export AI chats (ChatGPT, Claude, Gemini, DeepSeek) to professional PDF, Word, and Markdown documents. Save your history.
115
★ 1.0
GrabChat - Export & Backup AI Chats and save context in JSON, MD, and PDF
Export conversations from all major LLM platforms to JSON, Markdown, and PDF
341
★ 5.0
ChatShuttle | Sync & Restore AI Chats
Sync, backup, and search ChatGPT, Gemini, Claude chats in one place. Restore full context with images. Stored in your Google Drive.
36
★ 5.0
Data sourced from the Chrome Web Store · last verified Jun 10, 2026.