Skip to article frontmatterSkip to article content
Site not loading correctly?

This may be due to an incorrect BASE_URL configuration. See the MyST Documentation for reference.

Things on this page are fragmentary and immature notes/thoughts of the author. Please read with your own judgement!

AI Agent Code CLI

Most opensource AI agent code cli tools

Table 1:AI Agent Code CLIs Comparison

Agent CLI

LLM Agnostic?

Multi-Model?

Remote API?

Mobile App?

Primary Focus

Language

License

Popularity (Stars)

OpenCode

Y

Y

Y[1]

Y[2]

Full ecosystem (CLI, Desktop, Mobile, SDK)

TypeScript

MIT

~192.7k ★

Pi

Y

Y[3]

Y[4]

N

Minimalist, extensible harness

TypeScript

MIT

~82.8k ★

Goose

Y

Y

Y[5]

N[6]

General OS execution / testing

Rust / TS

Apache 2.0

~52.2k ★

Aider

Y[7]

Y[8]

Y[9]

N

Git-integrated pair programming

Python

Apache 2.0

~47.3k ★

Reasonix

N[10]

Y[11]

Y[12]

N[13]

DeepSeek prefix-cache optimization

Go / TS

MIT

~31k ★

Crush

Y[14]

Y[15]

N[16]

N

Elegant TUI with LSP integration

Go

MIT

~27k ★

Plandex

Y

Y[17]

Y[18]

N

Multi-step planning for large tasks

Go

MIT

~15.6k ★

Oh-My-Pi

Y

Y[19]

Unknown

N

Hash-anchored edits, subagents

Python / TS

Open Source

~11.6k ★

ForgeCode

Y[20]

Y[21]

Unknown

N

Multi-model MCP-driven coding

Rust

Apache 2.0

~7.5k ★

Pi Agent Rust

Y

Y[22]

Y[23]

N

High-perf, zero-unsafe isolation

Rust

Open Source

~1.5k ★

Codex CLI (OpenAI Official)

N[24]

N[25]

Y[26]

N

Enterprise proxy/sandbox workflows

Rust

Apache 2.0

N/A

Claude Code

N[27]

Y[28]

Y[29]

Y[30]

Native terminal integration w/ MCP

TypeScript

Closed

N/A

References

Footnotes
  1. JS/TS SDK.

  2. Dedicated OpenCode Mobile app.

  3. toolModel: separate reasoning vs. tool-calling model.

  4. RPC / SDK.

  5. REST API.

  6. Desktop & CLI only, no mobile app.

  7. 100+ providers via LiteLLM.

  8. --model/--editor-model flags.

  9. --message flag for headless/one-shot runs.

  10. DeepSeek models only.

  11. Executor + planner via planner_model.

  12. JSON-RPC interface.

  13. Desktop / VS Code integration only, no mobile app.

  14. Multi-model support across providers.

  15. Large model for the Coder agent, small model for the Task agent.

  16. Focuses on the TUI; no dedicated remote API.

  17. Per-role model packs: planner/architect/coder/builder/summarizer.

  18. Client/server architecture.

  19. Role-based routing, e.g. a dedicated plan role.

  20. 300+ models via OpenRouter and direct providers.

  21. Forge/Muse/Sage agents, each configurable with its own model.

  22. Ports Pi’s toolModel dual-model design.

  23. Built-in HTTP server.

  24. Locked to OpenAI models officially; community forks add other providers.

  25. Plan mode inherits the global model; manual switching only, no built-in per-role split.

  26. Python SDK.

  27. Anthropic models only (official).

  28. Per-subagent model configuration.

  29. Via MCP servers and hooks.

  30. Session handoff to mobile via QR code.