Phone and Mac, in step Open the chat with your Mac on your phone and you land in the conversation the Mac has open. Replies appear on every device as they are written. Messages that wait their turn A message for a conversation another app has open (an editor, a terminal) now waits and is sent as soon as that app lets go. Remote desktop See and control your own Mac from your phone. Organized conversations Groups by project, your own groups, Archived and Unread, synced across your devices. Full-text search across your conversations, Soul and Memory. Claude on the Agent SDK A new engine for agents that run on an Anthropic API key. API keys in one place Add your AI provider keys once on the website (Dashboard > Settings > Deploy Credentials); every agent picks them up. Plus many fixes to streaming, sync and reliability. Also in this build Sign in with Apple lands in the same account on every platform. Chat: Esc stops a reply, long messages and runs of tool calls fold, failed tools show why, file paths open the file, and the transcript no longer goes blank after a send.

Attach files to a chat Send a PDF, CSV, log or archive straight to your agent. The file lands on your own server, and the agent can open, search and run it for the rest of the conversation. Usage panel A new panel on the agents list shows what each model has cost you, broken down per model, plus the remaining balance on any prepaid provider account you have connected. Point an agent at your own model server Alongside renting a GPU, you can now give an agent the address of an OpenAI-compatible server you run yourself. The address is checked from the agent's machine before it is saved, so an unreachable one tells you immediately instead of failing later. Stays signed in Your agent now renews its own AI provider login, so a long run no longer stops because a token quietly expired while you were away. Notifications that work Fixed the cause of missing "your agent finished" notifications. Smaller fixes Clearer error messages throughout, and the chat now says when the conversation has been compacted.

Completely rebuilt as a fast, fully native app. - Deploy AI coding agents to your own server in minutes and chat with them from anywhere - Chat history now syncs across all your devices - Improved sign-in: Apple, Google, email, and phone - Delete Account is available in the Account tab - Many bug fixes and performance improvements

What's new in 2.1: • Per-Agent AI Model — Assign Claude, GPT-5, Gemini, or Grok individually to each agent • Real-Time Deploy Logs — Watch your server come online step by step • Improved Multi-Cloud Dashboard — Clearer agent status across all providers • Bug fixes and performance improvements

What's New in AgentAura v2.0.45 Major Update: One-Tap Cloud Deployment This release transforms the app into AgentAura — your mobile cloud deployment hub. Deploy AI-powered servers to major cloud providers directly from your iPhone. New Features: • One-Click Cloud Deployment — Launch a fully configured cloud instance in under 60 seconds. No terminal, no SSH, no complexity. • Multi-Cloud Provider Support — Deploy to Vultr (available now), with DigitalOcean, Hetzner, Akamai (Linode), Cloudflare, Railway, Google Cloud, AWS, Azure, and Oracle Cloud coming soon. • Flexible Server Plans — Choose from multiple configurations: 1 CPU/2GB RAM ($10/mo), 2 CPU/4GB RAM ($20/mo), 4 CPU/8GB RAM ($40/mo), 6 CPU/16GB RAM ($80/mo), and more. • Instance Naming — Personalize each deployment with custom instance names for easy identification and management. • Cloud Provider Dashboard — View and manage all your cloud instances from a single, intuitive interface. • Redesigned UI — Completely refreshed dark-mode interface optimized for cloud deployment workflows. • Enhanced Security — API keys stored securely on-device. All connections encrypted end-to-end. • Performance Improvements — Faster app launch, smoother navigation, and reduced memory usage. • Bug Fixes — Resolved various stability issues and improved overall reliability. Deploy your own cloud server today — it's never been this easy.

1. Secure, free, and fast internet access proxy aggregation. 2. large-scale chat model aggregation. 3. secure data retrieval aggregation.