Build a real, usable workbench with CloudBase from 0 to 1, then grow it with media sync, per-user permissions, shareable links, scheduled jobs, and payments.
GLM 5.3 is live on CloudBase: +50% coding capability, ship right after writing, and switch models by changing one parameter.
CloudBase now supports PostgreSQL — standard SQL table creation, auto-generated APIs, second-level Database Branch cloning, idle compute suspension, and high Supabase compatibility. Tencent's own agents already run on it.
Tusi uses CloudBase to get platform-grade multi-tenancy, built-in auth and production hosting in one go, generating an independent app for every user.
OpenAgentKernel is open source: an agent framework natively built for CloudBase, with session persistence, tool orchestration and human-in-the-loop built in.
The CloudBase connector is now on Zhipu ZCode: install one plugin and operate your entire CloudBase backend in natural language.
CloudBase Plugin is here — query the database, create cloud functions, and deploy from chat in Codex, Claude Code, and WorkBuddy. One prompt to connect; MCP, Skills, and Hooks bundled.
Tencent Cloud CloudBase at WAIC 2026: in the agent era, the cloud must evolve from a human-facing interface into a production line for agents.
Kimi K3 is live on CloudBase: 2.8 trillion parameters, million-token context, #1 on the frontend coding blind test, three ways to plug in.
Meiya Fengniao uses CodeBuddy + CloudBase to push OPC, letting non-technical colleagues deliver apps independently.
WeChat DevTools opened the whole mini program cockpit to external AI agents, and CloudBase completes the cloud loop — one sentence from idea to real-device preview.
CloudBase billing upgrade: resource points become the unified currency, with cloud resources and AI tokens charged from the same pool — one plan for both.
A Kujiale business-role staffer uses VS Code + Codex + CloudBase to ship a backend-and-database custom app in three to four days — one person absorbing the long-tail customization demand.
The mini program full-stack plan: tokens and cloud resources in one, AI coding tools plug straight in, ship right after writing code.
GLM 5.2 is now live on CloudBase: a Token Plan that comes with cloud resources, so you can ship code right after writing it. HY3, GLM 5.2, DeepSeek, Kimi and Qwen all in one pool.