APUS AI has open-sourced its Personal Agent product, Reflex Agent, built on its self-developed Reflex decision model and Continuum memory engine. This agent aims for more efficient decision-making and long-term task execution, reportedly achieving response speeds up to three times faster than models like Claude Code and Codex in tests. Reflex Agent utilizes a 'Loop Working' mechanism, where the Reflex model handles routing, skill selection, and supervision, driving tasks through a continuous cycle of decision, execution, and feedback. Its Continuum memory engine supports a form of 'infinite context' by continuously accumulating task information and preserving original content for decision-making, reducing reliance on summarized histories. AI
IMPACT This release could offer developers a new agent framework with potentially faster response times and enhanced memory capabilities for long-running tasks.
RANK_REASON This is a product release from a company that is not a frontier AI lab, focused on an agent framework.
AI-generated summary · Google Gemini · from 1 sources. How we write summaries →
