Zedra vs AI

Your coding agents run on your own machine. Zedra puts them in your pocket. Start a task on your laptop, walk away, and keep working from your phone. Zedra connects straight to your development machine, so you can watch agents work, answer their questions, review the diff, and start the next task from anywhere. WORKS WITH THE AGENTS YOU ALREADY USE Claude Code, Codex, opencode, and any other CLI agent or shell command. Zedra runs a real terminal, not a chat mirror, so nothing about your setup has to change. RUN A SWARM, NOT A SESSION Most work is several agents at once. Zedra shows every session on one screen with live status, so you can see which agent is thinking, which one finished, and which one is blocked waiting on you. NEVER MISS AN APPROVAL Agents stop and ask before they touch anything important. Zedra sends a push notification the moment one needs you, so a waiting agent is never a surprise you find twenty minutes later. A REAL TERMINAL Full terminal emulation with color, mouse reporting, and TUI support. A keyboard row built for code gives you arrows, tab, escape, and control without fighting the on-screen keyboard. Paste screenshots straight into a prompt. READ THE DIFF BEFORE YOU MERGE Syntax-highlighted diffs, file trees, and git status. Select code in the editor or a file preview and send it straight to an agent with Add to Chat, without switching apps. CONNECTS DIRECTLY, NOT THROUGH US Pair by scanning a QR code once. Zedra builds an encrypted peer-to-peer connection between your phone and your machine, and falls back to a relay only when your network blocks a direct route. Relays forward encrypted packets they cannot read, so Zedra never sees your code. Agents still talk to whichever provider you configured them with. OPEN HOST The host daemon is open source and runs on macOS and Linux. Install it with one command, keep it running in the background, and pair as many devices as you want. Requires a Mac or Linux machine running the Zedra host.

AI-IDE is a lightweight editor for iPhone, iPad, and Mac. Edit Markdown, code, and config files inside iCloud Drive or any Files folder — and work with the LLM of your choice without leaving the document. KEY FEATURES • Mark up the preview, let AI edit — Tap a spot or select a passage and say what to change. Queue marks, send one batch, review one diff. • Bring your own AI — Anthropic, Gemini, DeepSeek, OpenRouter, Groq, Mistral, any API-compatible endpoint, local Ollama, or an AISA key for 70+ models. 16 one-tap presets. No built-in account or key: nothing runs until you add one. • You approve every change — Edits show a diff first; nothing is written until you tap Apply, and overwrites are snapshotted. • Multi-tab editor — Syntax highlighting for 10 languages, line numbers, autosave, find & replace. • Rich Markdown preview — Outline navigation, in-document search, reading progress with resume, code copy, task checkboxes, light/dark themes. • Office files and PDFs, both ways — Export to PDF, Word, Excel, and PowerPoint on device; "AI Refine" turns a document into slides, using your own .pptx as the master and Brand Kits for colours, fonts, and logo. Open a .docx, .xlsx, .pptx, or PDF to edit in place, formatting kept. • AI illustrations — One prompt draws an illustration at the cursor with your own image model; it flows into PDF, Word, and PowerPoint. Decks can auto-illustrate. • Built-in data tools — 35 the AI calls itself: web pages, academic papers, US financials and SEC filings, crypto prices, X posts, prediction markets, Perplexity. Switched on by category, with the token cost shown. HTTPS-only and host-whitelisted; your key is injected on device, never seen by the model. • Prompt Library — Activate several presets at once, merged in your order with a size meter. Three research procedures are built in; add more from the library or GitHub. • KB Library — File a document once and reference it as often as you like: type ~ to attach one, or let the AI search the library and read only the passage it needs. Ranked search working in Chinese and says so when nothing matches. • Share what you tune — Prompts, export styles and KB entries export as files. Install from the built-in library or add a source: any public GitHub repo or page. Updates are flagged, never automatic, and an edited copy is never overwritten. • Tool-using AI — Reads, searches, and writes workspace files. Optional web search via Brave, Tavily, Perplexity, or AISA. • Files done right — Open files from anywhere; create, rename, move, duplicate, delete; multiple pinned workspaces. • Local version control — Snapshot, diff, and revert on device: no git account, history never leaves it, instant in huge folders. Auto-commit optional. • Reliable by default — Suggests the right model for the job; dropped streams resume without re-sending the conversation. • Multimodal — Attach files, photos, or clipboard images. Vision models get them directly; PDF / DOCX / PPTX / XLSX are transcribed. • Trilingual — Full UI in English, 繁體中文, 简体中文; switch without a restart. • iPad and Mac — Three-pane Split View on iPad and Mac, tabs on iPhone, a full Mac menu bar and keyboard shortcuts. PRIVACY • API keys live in the Keychain (device-only). Your content goes straight to the endpoints you configure — no proxy, no telemetry. Chat history is local. REQUIREMENTS • iOS / iPadOS 17.0 or later, or macOS 14.0 or later. You provide the API keys for the services you use. For support, send mail to sportusagi@gmail.com. AI-IDE Pro is an auto-renewable subscription. Payment is charged to your Apple ID account at confirmation of purchase. The subscription renews automatically unless cancelled at least 24 hours before the end of the current period; manage or cancel it in your Apple ID account settings. Terms of Use (EULA): https://www.apple.com/legal/internet-services/itunes/dev/stdeula/ Privacy Policy: https://www.termsfeed.com/live/92fad2c6-6901-4de6-9110-e4aa7aa1f7d8