EdgeOne Makers

EdgeOne Makers is a platform for deploying web apps and AI agents on Tencent EdgeOne infrastructure. It combines hosting, functions, storage, runtime, sandbox tools, memory, observability, and built-in models.

AI 에이전트 개발 AI 개발 도구 AI 앱 빌더

웹사이트 방문

Overview

EdgeOne Makers is a development and deployment platform built on Tencent EdgeOne infrastructure. The product description says it provides serverless deployment for frontend pages, dynamic APIs, and AI applications on the same platform.

The site positions Makers for teams building modern web projects and AI agents. The home page highlights web app hosting with functions and storage, while the agent documentation adds runtime, sandbox tooling, conversation memory, observability, and built-in models for agent workloads.

Core capabilities

Multiple project types

Deploy static sites, single-page apps, full-stack apps, and agent apps from the same platform, using Git-based workflows, templates, or direct upload.

Agent runtime

Use managed runtime for agent loops, LLM calls, orchestration, and session-based execution so long-running conversational tasks can keep state.

Sandbox tools

Run browser, code execution, shell, and file operations inside an isolated sandbox for agent workflows that need tool use or file handling.

Conversation storage

Store sessions and messages with a universal conversation storage API that supports memory management across agent frameworks.

Observability

Collect traces with zero-intrusion instrumentation and view them locally or in the cloud for debugging and monitoring agent behavior.

Built-in models

Use built-in models with automatically injected gateway keys, plus a limited-time monthly trial quota for new users.

Common use cases

Modern web app deployment
Ship marketing sites, SPAs, or full-stack web apps with serverless deployment, hosting, functions, and storage.
Agent applications
Build conversational assistants and other multi-turn agents that need session stickiness, memory, and tool use.
Tool-heavy workflows
Run code-using or file-processing agents that work with browser actions, shell access, and isolated file operations inside a sandbox.
Mixed runtime projects
Separate long-running agent orchestration from stateless business logic by combining agents/ and cloud-functions/ in one project.
Model-enabled prototypes
Create AI applications with built-in model access and a gateway-backed key injection flow instead of wiring model access manually.

Pros and Cons

Pros

Combines web app deployment and AI agent development in one platform.
Supports Git-based deployment workflows and multiple ways to start a project, including templates and direct upload.
Includes built-in agent infrastructure such as runtime, sandbox tools, memory, tracing, and models.
Offers a permanently available free plan according to the pricing page.
Keeps web and agent projects under shared management for accounts, deployments, monitoring, and domains.

Cons

The source does not provide detailed paid-plan pricing or quota numbers beyond the free tier and contact-sales flow.
Some capabilities are described at a platform level rather than with step-by-step product limits, so readers may need the docs to confirm implementation details.

FAQ

What does EdgeOne Makers support deploying?

EdgeOne Makers supports deploying web apps and AI agents from one platform. The pricing page says it includes web app deployment, AI agents, built-in models, cloud functions, edge functions, and blob/KV storage, while the agent docs add runtime, sandbox tools, conversation storage, observability, and built-in models for agent projects.

Is there a free plan?

The pricing page states that the Free plan is permanently available. It also says more individual and enterprise pricing plans are coming soon, and that customized services can be requested by contacting the team directly.

What kinds of agent projects can it handle?

The agent documentation says you can choose frameworks such as Claude, OpenAI, LangGraph, and CrewAI, with no language restriction between JavaScript and Python. The same docs describe both agents/ runtime for conversational workloads and cloud-functions/ for stateless business logic.

Can web apps and agents live in the same project?

Yes. The platform documentation says Web and Agent projects share the same project, with accounts, deployments, monitoring, and domains centrally managed.

What happens if I exceed the free-tier usage limits?

The pricing FAQ says that before the paid version launches, service will not be interrupted if free-tier usage limits are exceeded, though abnormal traffic and abusive behavior may still be monitored.