Synara v0.8.2: GPT-6 Astra available with a ChatGPT subscription
Synara v0.8.2 lets users access GPT-6 Astra through their ChatGPT subscription, alongside big speedups in rendering, diff sorting and file tree building.
Synara v0.8.2 is out 🪽 ✨ GPT-6 Astra is available with your ChatGPT subscription ⚡ Smoother streaming with 31.6% less renderer CPU and frame times dropping from 25.0 ms to 9.6 ms 🧮 Large diff sorting is 92.9% faster and file tree building is 89.9% faster 📝 Rename chats instantly with /rename or let Synara generate the title ➕ Add selected text from Markdown previews directly to chat 🚀 Large projects start faster with less replay work and smarter memory usage 🔎 Faster, more reliable model discovery across providers 🤖 Pi now supports authenticated OpenRouter models and updated OpenCode Zen models 🎯 Claude Auto compaction now lets Claude Code choose the right context window 🧩 Claude context-window information is now available to MCP agents 📊 Droid usage limits now work with Factory credentials 🛠️ Devin sessions recover more reliably when tools go quiet or child processes stall 🔁 Queued messages dispatch correctly and no longer duplicate streamed replies 📜 Scrolling stays under your control while tools run and replies stream 🧱 Streaming Markdown keeps code blocks stable without unnecessary remounts 🗂️ File, image, PDF, and diff previews update when files change without overwriting edits 🧵 Long messages preserve text and Unicode correctly across pagination and handoffs 🧾 Final answers stay complete after tools, folded output, and segmented replies ✅ Read tasks stay read and approval cards recover correctly after restart 🔁 Kilo Code has been migrated to OpenCode, including sessions, tasks, favorites, and editor tabs 🪟 Better Windows provider detection and safer process handling ⚙️ TypeScript 7, Bun 1.4.2, faster checks, and parallel CI lanes ✨ A major performance, reliability, and everyday workflow update across Synara