Qingcheng Jizhi pitches a full-stack Token infrastructure for the agent economy
新智元 · wechat · 2026-07-21
The article describes Qingcheng Jizhi’s “front shop, back factory” Token infrastructure stack for the agent era.
Three layers
- Back factory: token production with a custom inference engine called Chitu plus the Bagua Furnace software stack.
- Middle layer: AIPing, a neutral token routing and benchmarking platform that monitors dozens of providers and hundreds of model endpoints around the clock.
- Front layer: AgentStore and ATM (AgentTokenManager) for agent distribution, metering, cost allocation, and resource isolation.
Key claims
- The stack is designed to solve unstable APIs, poor portability of foreign inference engines on domestic GPUs, and opaque multi-agent costs.
- The company emphasizes neutrality across heterogeneous Chinese accelerators.
- It also argues that idle domestic compute could be turned into exportable token supply for overseas developers.
The piece presents this as a full-stack infrastructure play for China’s AI agent ecosystem, not just a single product.
More from coding & agent
- A roundup of AI agents and MCP resources, including how to evaluate agents — _jaydeepkarale · 2026-07-21
- A full course shows how to build and deploy an AI agent with OpenAI and LangChain — _jaydeepkarale · 2026-07-21
- A beginner guide to AI agents points readers to a Stanford webinar — _jaydeepkarale · 2026-07-21
- A practical guide on how to evaluate AI agents — _jaydeepkarale · 2026-07-21
- MCP is headed toward easier scale, event-driven extensions, and workable file uploads — EricBuess · 2026-07-21
- Developers debate the missing composition model for AI agents — threepointone · 2026-07-21