Zed

Zed is GPL-licensed Rust editor whose agent panel can host Claude Code, Codex and Gemini CLI. We compare it in Coding agents. It is 2nd of 9 on GitHub stars (62,000). The trade-off is being marked No on Headless / CI, where Claude Code records Yes.

What is Zed?

Zed — GPL-licensed Rust editor whose agent panel can host Claude Code, Codex and Gemini CLI. Zed is a fast, open-source Rust editor with a native agent panel, but its most interesting property in this category is the Agent Client Protocol: external agents run inside Zed's review surface, so you can drive Claude Code or Gemini CLI from an editor rather than a terminal. Bring your own key for free, or pay for hosted models. Local models work. There is no sandbox and no pull-request path. Founded 2021. Open source under GPL-3.0.

We track it in 1 comparison — Coding agents — so every claim below is a cell in a table you can open and check rather than an impression. Across those rosters it sits against 16 other tools, and what follows is where it visibly separates from them.

Where it wins.

  • GitHub stars — 62,000, 2nd of 9.

Each of those is ranked against the whole roster on its category page, not against a hand-picked subset, so a first place here means first of everything we list.

Where it gives ground.

  • Headless / CI — No. Claude Code records Yes.
  • Subagents — No. Claude Code records Yes.
  • Background agents — No. Claude Code records Yes.
  • Hooks — No. Claude Code records Yes.

None of these disqualify it on their own. They are the fields to check against your own requirements before you commit, because they are the ones where a competitor genuinely does better.

Provenance. 24 of 26 tracked fields carry a value for Zed, and 4 of those cite a document you can open. Last verified 2026-01-15. Every figure keeps its own provenance — measured by us, claimed by the vendor, inferred, or community-reported — and we would rather print a dash than a guess.

Its nearest neighbour in our data is Aider. Zed is ahead on MCP (Yes against No) and Permissions (granular allowlist against prompts every action). Aider takes Headless / CI (Yes against No) and Licence (Apache-2.0 against GPL-3.0). That pattern repeats across the rest of the roster — see Zed alternatives for the other rivals, each compared the same way.

At a glance
Founded
2021
Licence
GPL-3.0 (open source)
Fields we track
24 of 26
Last verified
2026-01-15

Zed in Coding agents

Ranked against 17 tools across 26 sourced fields. Open the full Coding agents table.

The hedge in this category: a fast GPL editor that hosts other people's agents over ACP, so your harness choice stays reversible while you keep a proper review surface. Weak as an agent in its own right — a read-only Ask profile rather than a real plan-and-approve gate, no subagents, no PR path, no headless mode.

Where it lands in this roster
Headless / CI
No15th of 16
Inferred
Subagents
No10th of 14
Inferred
Background agents
No12th of 16
Inferred
Hooks
No7th of 10
Inferred
Licence
GPL-3.09th of 15
Vendor-claimedverified 2026-01-15source
GitHub stars
62,0002nd of 9
Community-reported
Opens PRs
No10th of 16
Inferred
BYOK
Yes1st of 16
Inferredverified 2026-01-15

When to use Zed

Zed is the right call in these situations, each one drawn from a field we actually record:

  • Driving Claude Code or Gemini CLI from an editor instead of a terminal.
  • Teams who want a genuinely open-source editor with local-model support.
  • GitHub stars is your binding constraint. Zed records 62,000, 2nd of 9 in the Coding agents roster. We define that field as stars on the primary open-source repository, rounded to the nearest thousand.

When not to use Zed

Reach for something else when any of the following is a requirement rather than a nice-to-have:

  • Headless / CI. Zed records No on the Coding agents table. Claude Code records Yes on the same field. If that is a hard requirement rather than a preference, start elsewhere.
  • Subagents. Zed records No on the Coding agents table. Claude Code records Yes on the same field. If that is a hard requirement rather than a preference, start elsewhere.
  • Background agents. Zed records No on the Coding agents table. Claude Code records Yes on the same field. If that is a hard requirement rather than a preference, start elsewhere.
  • Hooks. Zed records No on the Coding agents table. Claude Code records Yes on the same field. If that is a hard requirement rather than a preference, start elsewhere.
  • Opens PRs. Zed records No on the Coding agents table. Claude Code records Yes on the same field. If that is a hard requirement rather than a preference, start elsewhere.

We publish this block because a comparison that only lists what a tool is good at is marketing. Every figure above sits on the same page as its source, and the field definitions are on the category tables if you want to check how we measured them.

Tools compared alongside Zed

Everything below shares at least one comparison category with Zed, ordered by how much overlap there is. For the reasoning on each — which fields it wins, which it loses — see Zed alternatives.
Minimal terminal pair-programmer with a tree-sitter repo map and automatic git commits.
Sourcegraph's opinionated agent with no model picker and an ad-supported free tier.
Open-source VS Code and JetBrains agent built around an explicit plan/act split.
Cursor
VS Code fork with a real codebase index, background cloud agents and a CLI.
Fully delegated cloud engineer: file a task in Slack or the web app, get a PR back.
DIY harnessbaseline
Your own loop over a model API with an agent SDK — the build-it-yourself reference point.

Recent Zed changes

We have not logged a dated change for Zed yet. The timeline fills in as pricing moves, features ship and things get deprecated.

Sources and gaps

What we don't know. 2 of the 26 fields we track for Zed are still blank: Terminal-Bench and Bench model. Those render as dashes rather than as zeroes or assumptions, because an empty cell and a bad cell are not the same thing and only one of them is honest. If you know any of these figures and can point at a document, tell us.

Every figure on this page traces back to a document you can open. Where a vendor claims a number we could not reproduce, the cell says so.