Loopa is an AI agent aimed at complex multi-step tasks: summarising and analysing PDFs, building presentations, generating research reports, automating email replies, running data and financial analysis, and creating quizzes, all from one interface. It routes across multiple providers including Qwen, StepFun, DeepSeek, Gemini, Hailuo AI, GPT, Claude, GPT Image 2 and Seedance 2.0.
The breadth is organised around real job categories rather than being a bare model picker: document analysis, presentation building and financial reporting are genuinely different workflows, and having each addressed as its own capability rather than requiring the user to prompt-engineer a general chat box into doing each one is a meaningfully more usable product. Email auto-replies specifically are worth using carefully – review before anything goes out under your name automatically.
Routing across nine-plus underlying providers means output consistency depends on decisions Loopa makes about which model handles which request type, not something visible or controllable from outside. No pricing figures are shown beyond a free starting option, so the actual cost of the financial reporting and research-report features – likely the higher-value, higher-cost workflows – is not clear upfront.
