MicroManus

An autonomous deep-research agent with live internet access and usage-based billing. Bring your own LLM API key, ask a question, and MicroManus plans, searches, reads, and reasons in an agentic loop (think → call a tool → observe → think again) before handing you a cited answer or a downloadable PDF report. It ships a streaming multi-turn chat UI with a collapsible timeline of every search and read step, per-chat token and cost tracking priced per model, and a GitHub-OAuth paywall unlocked via coupon or a real Razorpay test-mode card payment. Built on Next.js 16, React 19, Tailwind CSS 4, and Supabase (Postgres, Auth, and row-level security), with support for Anthropic, OpenAI, and Moonshot (Kimi) models through OpenRouter, direct provider keys, or any OpenAI-compatible endpoint.