Claude Code alternatives: 11 coding agents and who each one suits

The closest Claude Code alternatives are Codex if you already pay for ChatGPT, Cursor if you want an AI editor with many models, GitHub Copilot for the lowest price and GitHub-native agents, OpenCode for an open source agent that works with any provider, and Google Antigravity for a free start with Gemini. Pick by two things: where you want to work (terminal, editor or cloud) and which model subscription you already have.

The alternatives at a glance

Prices are the cheapest paid tier as of October 2026, from each product's own pricing page.

The 11 alternatives

1. Codex

OpenAI's coding agent comes with every ChatGPT plan. Free and Go users get it in the ChatGPT desktop app; Plus ($20) adds the CLI, the IDE extension (VS Code, Cursor, Windsurf) and cloud tasks on the web, GitHub and Slack. The CLI is open source under Apache-2.0. It reads AGENTS.md, Agent Skills, MCP servers and hooks, so a Claude Code setup ports over with little work. Suits: anyone who already pays for ChatGPT. Full comparison: Codex vs Claude Code.

2. Cursor

An AI editor with a CLI and cloud agents. Since October 2026 its plans are part of The X Subscription: Free, Plus $8, Premium $30, Super $100, Ultra $200. Cursor's own Grok and Composer models are in the plan; Claude, GPT and Gemini cost API list price on Premium and up. Automations run cloud agents on a cron schedule or on GitHub and Slack events. Suits: people who want to see every diff in an IDE and switch models. See Cursor vs Claude Code.

3. GitHub Copilot

The cheapest way into a full agent: Free, Pro $10, Pro+ $39, Max $100, Business $19 and Enterprise $39 per user. Its cloud agent takes an issue assigned to @copilot, works in a GitHub Actions environment and opens a pull request. Copilot CLI supports skills, custom agents, hooks, MCP and plugins. Every plan lists Claude, GPT, Gemini and Grok models, and Pro+ and Max can run Claude and Codex as third-party agents (preview). Suits: teams that live in GitHub. See Copilot vs Claude Code.

4. OpenCode

An open source (MIT) terminal agent, with a desktop app in beta and an IDE extension, that connects to 75+ model providers, including local models. You bring a key, pay per token through OpenCode Zen, or subscribe to OpenCode Go ($10 a month) for open coding models. It has a read-only plan agent, Agent Skills, MCP and JavaScript plugins. Note that Anthropic does not allow Claude Pro or Max subscriptions in third-party tools, so Claude here means an API key. Suits: open source fans and anyone using several providers. See OpenCode vs Claude Code.

5. Google Antigravity

Google's agentic development platform: an App for running several local agents in parallel, the agy CLI, a full IDE with an agent manager, and an SDK, on macOS, Windows and Linux. The individual plan is free with weekly rate limits; Google AI Pro and Ultra raise them and add Gemini 4 Argon. Claude Sonnet 5.5 and Opus 5.5 are also on the model list. It supports MCP, Agent Skills, lifecycle hooks, subagents, plugins and a /schedule command. Suits: a free start, or Gemini users.

6. Gemini CLI

Open source (Apache 2.0) and once the go-to free terminal agent, but since June 18, 2026 it no longer serves personal Google accounts or Google AI Pro and Ultra, which moved to Antigravity CLI. It still works with Gemini Code Assist Standard (1,500 requests a day) and Enterprise (2,000) and with paid API keys. Suits: companies with Code Assist licenses.

7. Kiro

AWS's agent built around spec-driven development: it turns a prompt into requirements, a design and a task list before writing code. It has an IDE, a CLI, a web agent that opens pull requests, and Kiro Crew for always-on personal tasks. Pricing is credit-based: Free (50 credits), Pro $20 (1,000), Pro+ $40, Pro Max $100, Power $200, with Claude Sonnet 5 and Opus 5 among the premium models. Suits: teams that want structure before code. See Kiro vs Claude Code.

8. Cline

An open source coding agent for VS Code and the terminal. There are no subscriptions or seat fees for the open source version: you pay for inference at cost or bring your own OpenAI, Anthropic or Google key. The Enterprise plan adds JetBrains, SSO and team controls. Suits: developers who want an agent in VS Code and full control of model spend.

9. Windsurf

An AI editor with its own agent, Cascade, and its own SWE-1.5 model alongside the major providers. Free, Pro $15 a month with 500 prompt credits, Teams $30 per user. Suits: people who want an editor-first agent at a lower price than most.

10. Amp

A coding agent you run from the terminal (amp), on servers and CI (amp --no-tui), or on orbs, remote machines that keep working after you close your laptop and that teammates can join. The Hobby tier is free with pay-as-you-go orb time; Individual is $20 a month, and Amp says you can use your ChatGPT, Claude and other subscriptions. Suits: people who want hosted, shareable agent machines.

11. Aider

The original terminal pair programmer: open source (Apache-2.0), works with almost any model including local ones, and commits each change to git with a sensible message so you can review or undo it. Free; you pay your model provider. Suits: developers who want a small, git-first tool.

How to choose

• You already pay for ChatGPT: Codex. For Claude: stay with Claude Code. For Google AI Pro: Antigravity.
• You want an editor, not a terminal: Cursor, Windsurf, or Copilot in VS Code.
• You want open source: OpenCode, Cline or Aider, or the Codex CLI.
• You want the lowest monthly bill: Copilot Pro at $10, or an open source agent with a cheap model.
• You want structure and specs: Kiro.

You don't have to settle on one. Many developers keep two agents for different jobs, and the same Agent Skills and MCP servers work across most of them. Briefkin's installer hooks into Claude Code, Codex, Cursor, OpenCode, Gemini CLI, Copilot CLI, Windsurf and Amp, so all of their sessions land on one board and you get a ping when one needs you.