Home
MiniMax Unveils MaxHermes, World's First Cloud Sandbox for Autonomous AI Skill Extraction and Evolution
On April 16, MiniMax (00100.HK) unveiled Hermes — the world's first cloud-based sandbox, MaxHermes. This assistant is powered by the Hermes Agent, with its key innovation being a "learning loop" mechanism that lets it autonomously extract reusable "Skills" from complex tasks and save them as standalone documents. Unlike tools like OpenClaw, which depend on manually predefined capabilities, MaxHermes' skill library grows dynamically through use and self-iterates based on user feedback.

From a technical standpoint, MaxHermes combines persistent cross-session memory, natural language-defined scheduled tasks, and parallel operation of multiple sub-agents to create an evolutionary intelligent agent capable of long-term operation. Currently, this assistant primarily runs on MiniMax's latest M2.7 programming model. This model shows significant improvements in tool call accuracy, adherence to complex instructions, and compatibility with Agent Harness, making it one of the most active models in the Hermes ecosystem. It provides the foundational computing power for the autonomous generation and precise execution of Skills.
The launch of MaxHermes signals a shift for AI Agents from "fixed capability tools" to "self-evolving entities." By combining the logic of learning loops with model iteration, MaxHermes significantly expands AI's ability to handle ambiguous and long-term tasks. Its continuous alignment with user preferences over time is expected to boost the practical adoption and quality of AI in personal assistant and enterprise automation scenarios.
Related article
Inside Details Exposed About Next-Gen Gemini: Strained Computing Power, Internal Teams Disagreed on Development Priorities and Resource Allocation
Reports indicate that the launch of Google’s highly anticipated next-generation Gemini model has been pushed back. Internal disagreements over development priorities and resource allocation, combined with limited computing capacity and complex approv
OpenAI Dismisses Growth Slowdown Concerns, Says Multiple Business Units Accelerating
In response to external scrutiny regarding decelerating sales growth and missed internal benchmarks, AI leader OpenAI issued a confident statement on Tuesday, April 28. The company clarified that its consumer products and enterprise services are adva
Alibaba Super Cup: Qwen3.8-Max Debuts with Boosted Coding and Office Tools
Alibaba has officially unveiled Qwen3.8-Max, a next-generation foundation large model boasting 2.4 trillion parameters. This significant AI advancement delivers substantial performance gains in core areas like coding and professional office tasks, sh
Related Special Topic Recommendations
Comments (0)
0/500
On April 16, MiniMax (00100.HK) unveiled Hermes — the world's first cloud-based sandbox, MaxHermes. This assistant is powered by the Hermes Agent, with its key innovation being a "learning loop" mechanism that lets it autonomously extract reusable "Skills" from complex tasks and save them as standalone documents. Unlike tools like OpenClaw, which depend on manually predefined capabilities, MaxHermes' skill library grows dynamically through use and self-iterates based on user feedback.

From a technical standpoint, MaxHermes combines persistent cross-session memory, natural language-defined scheduled tasks, and parallel operation of multiple sub-agents to create an evolutionary intelligent agent capable of long-term operation. Currently, this assistant primarily runs on MiniMax's latest M2.7 programming model. This model shows significant improvements in tool call accuracy, adherence to complex instructions, and compatibility with Agent Harness, making it one of the most active models in the Hermes ecosystem. It provides the foundational computing power for the autonomous generation and precise execution of Skills.
The launch of MaxHermes signals a shift for AI Agents from "fixed capability tools" to "self-evolving entities." By combining the logic of learning loops with model iteration, MaxHermes significantly expands AI's ability to handle ambiguous and long-term tasks. Its continuous alignment with user preferences over time is expected to boost the practical adoption and quality of AI in personal assistant and enterprise automation scenarios.
Inside Details Exposed About Next-Gen Gemini: Strained Computing Power, Internal Teams Disagreed on Development Priorities and Resource Allocation
Reports indicate that the launch of Google’s highly anticipated next-generation Gemini model has been pushed back. Internal disagreements over development priorities and resource allocation, combined with limited computing capacity and complex approv
OpenAI Dismisses Growth Slowdown Concerns, Says Multiple Business Units Accelerating
In response to external scrutiny regarding decelerating sales growth and missed internal benchmarks, AI leader OpenAI issued a confident statement on Tuesday, April 28. The company clarified that its consumer products and enterprise services are adva
Alibaba Super Cup: Qwen3.8-Max Debuts with Boosted Coding and Office Tools
Alibaba has officially unveiled Qwen3.8-Max, a next-generation foundation large model boasting 2.4 trillion parameters. This significant AI advancement delivers substantial performance gains in core areas like coding and professional office tasks, sh











