Zed

Overview

Zed is a code editor written from scratch in Rust by the team behind Atom and Tree-sitter, built to use several CPU cores and the GPU so that typing, search and navigation stay responsive in large projects. The AI lives inside the editor rather than beside it: an Agent Panel prompts agents that read the project, edit across files and run tools, several threads can run in parallel across projects and worktrees, an inline assistant rewrites a selection in place, and Edit Prediction proposes the next change as you type from Zed's own open-source, open-data Zeta2 model. Models can come from Zed's hosted service, your own provider API keys, an existing subscription, a gateway or a local runtime, and MCP servers can be attached to give the agents extra tools. The multiplayer features Zed started with are still there, shared editing, screen sharing and project chat with teammates. It runs natively on macOS, Linux and Windows, and the source is public, licensed primarily under GPL-3.0-or-later with Apache-2.0 components where marked.

Best for developers who want a fast, native, open-source editor with agentic AI and the option to bring their own model keys.

Zed suits developers who care how an editor feels under their hands and want AI they can turn up or down rather than a chat window bolted on. Adopting it is a switch and not an addition, so it rewards anyone willing to rebuild their setup and frustrates anyone whose day hangs on one specific plugin from another IDE. Pointing it at your own provider keys keeps the AI spend predictable, which makes it an unusually cheap tool to evaluate properly. For a team standardised on a JetBrains or VS Code toolchain, the pull has to come from the editing experience itself.

Key Facts

Features

Where it runs

Its own native editor for macOS, Linux and Windows

Inline completions

Yes, edit prediction, with 2,000 accepted predictions on the free plan and unlimited on Pro

Agent mode

Yes, an agentic mode for multi-file edits

Model choice

Yes, Zed-hosted models or your own API keys from Anthropic, OpenAI, Google, Amazon Bedrock and others

Data control

Bring your own API keys so requests go straight to your provider; Business adds organization data governance

Ecosystem fit

Native editor built for speed, with multiplayer collaboration

API

N/A

Integrations

Integrations coming soon.

Deployment and Support

Support

• Community forum

Onboarding

• Documentation and knowledge base

Company size

• Freelancers
• Small and mid-sized
• Enterprise

Suitability

The table combines the industries the vendor addresses with the company sizes the tool is aimed at. An empty cell means the combination is not stated, not that it is ruled out.

Pros and Cons

Pricing

• Personal

  $0

  Free forever. 2,000 accepted edit predictions, or unlimited with your own API keys.

• Pro

  $10

  Per month. Unlimited edit predictions and $5 of monthly AI token credits; hosted model usage billed by token beyond that.

• Business

  $30

  Per seat per month. Adds organization AI policies and data governance.

Best For

• Technology

Alternatives and Comparisons

What users say

Reviews

Share your experience

Write a Review