Cursor alternatives
If Cursor is not the right fit, these solve similar problems.
GitHub Copilot
The most widely deployed coding assistant, available in most major editors.
The safe institutional choice rather than the most capable one. It lives inside the editor you already use, it is the tool corporate security teams have most often already approved, and it does the everyday job — finishing the line, filling in the obvious — without asking you to change anything about how you work. Where it is behind is agentic, repository-wide work, which is exactly where the newer entrants concentrate.
from $10/mo
Claude
Assistant known for long-context reasoning and writing that needs less editing.
Strongest on long documents and on writing that has to sound like a person wrote it. The large context window is the practical differentiator — you can put a whole contract, a long transcript or several files in at once and ask about them together, which changes what kinds of work are possible rather than just making them faster. Less of an all-in-one than its main competitor.
from $20/mo
Claude Code
Terminal-based coding agent that runs commands and edits files directly.
An agent that works in the terminal rather than an editor, which suits people whose workflow is already command-line shaped. It reads the repository, makes changes across files and runs the tests, and the honest advantage over completion tools is that it closes the loop — it can check whether what it wrote actually works. The honest disadvantage is the same as every agent: it will occasionally report success on work that needs another look.
from $20/mo
Codex CLI
OpenAI's open-source terminal coding agent, rewritten in Rust.
The obvious choice if your team already pays for ChatGPT, because the agent comes with the plan rather than as another line item. The client being open source matters more than it sounds: you can read what it sends and pin a version. What you are buying is the model behind it, not the tool, and that is worth being clear about — switching away from OpenAI means switching tools too.
from $20/mo
Devin Desktop (Windsurf)
The renamed Windsurf editor, now combining a full IDE with a command center for coding agents.
Positioned between an editor and an agent: you get a normal editing experience plus a mode that will go away and carry out a multi-step change. That combination is genuinely useful, and it is also the category's most crowded position — the practical question is not whether it works but whether it works better than the tool you already have open.
from $20/mo
Gemini CLI
Google's open-source terminal agent, with the most usable free tier of the group.
The cheapest way to find out whether a terminal agent fits how you work, because the free tier is large enough for real tasks rather than a demo. Treat that as its main role: an evaluation path and a second opinion, rather than the tool you standardise a team on. Free quotas at this size have historically been the first thing to change, so do not build a workflow that breaks when it does.
Free tier available
opencode
Open-source terminal coding agent that runs on whichever model you point it at.
The right default when vendor independence is an actual requirement rather than a preference — you can move to a different model without changing how you work. The cost of that freedom is that nobody has tuned the prompts for the model you chose, so quality varies more between setups than it does with a first-party tool. Budget some time for configuration you would not spend elsewhere.
Free
Pi
A minimal terminal coding harness you extend in TypeScript instead of configuring.
Worth it when you intend to build on the agent rather than only use it — the four packages are meant to be imported, and that is a genuinely different offer from a CLI with a plugin folder. The flip side is stated plainly in its own README: there is no built-in permission system, so it runs with whatever access the shell that launched it has, and the project's answer is to containerise. That makes it a poor fit anywhere the sandbox has to come from the tool.
Free
herdr
A background server that owns your agents' terminals, and knows which one is blocked.
Worth installing at the point where you run more than one agent at a time, because that is the point where a multiplexer stops being enough: tmux can tell you a pane went quiet, but not whether it went quiet because it finished or because it is waiting for your approval. herdr's whole value is that one distinction. Read the supported-agents table before you assume it applies to you, though — for Claude Code, Codex and Cursor Agent CLI, state is inferred by matching rules against the visible screen, and only agents with lifecycle hooks or plugins report it authoritatively.
Free
Phind
Search engine tuned for programming questions, with runnable answers.
Answers developer questions with sources rather than writing code in your editor, which makes it a replacement for searching and reading threads rather than for a coding assistant. That is a narrower job than the agentic tools do, but it is a job people still spend real time on, and having citations you can open is the difference between an answer and a guess.
from $20/mo