# Four ways to build an agent

Compare the paths by who builds, runs, and decides the architecture.

## Build it yourself

Code-first. You own the full stack.

Write the agent yourself: model calls, tool wiring, memory, the orchestration loop, deployment, monitoring. Frameworks give you blocks; the architecture is yours.

- Builds: You
- Runs: You
- Decides: You

Read more: https://agentailor.com/paths/build-yourself.md

## Build it with a coding agent

You decide the architecture. The agent writes the code.

Agentic engineering, not vibe coding. You make the architectural calls before the agent touches the keyboard; Claude Code or Cursor handles implementation. You review the decisions, not just whether it compiles.

- Builds: A coding agent
- Runs: You
- Decides: You

Read more: https://agentailor.com/paths/coding-agent.md

## Deploy an open-source agent

It already exists. Clone, configure, run.

The OSS agent space has matured. When something already does 80% of what you need, deploying it on your own infrastructure beats rebuilding it. The question is no longer "does it exist?" but "which one is worth deploying, and why?"

- Builds: The community
- Runs: You
- Decides: Shared

Read more: https://agentailor.com/paths/open-source.md

## Use a managed service

Someone else runs the harness. You configure via API.

The path to watch most closely in 2026. Session persistence, the execution environment, retry logic — these are rarely your competitive advantage. A managed service wires them so you can spend engineering time on what makes your agent actually better.

- Builds: The vendor
- Runs: The vendor
- Decides: Shared

Read more: https://agentailor.com/paths/managed.md

Paths can combine. A team on a managed service can still build on top of it; an open-source project can be forked and maintained independently.

Find your starting point: https://agentailor.com/#fitting

Use with AI: https://agentailor.com/for-agents

---
Source: https://agentailor.com/paths