Get AI summaries of any video or article — Sign up free
There’s a fast new code editor in town thumbnail

There’s a fast new code editor in town

Fireship·
4 min read

Based on Fireship's video on YouTube. If you like this content, support the original creators by watching, liking and subscribing to their content.

TL;DR

Zed is an open-source, Rust-built editor designed for very low latency and high responsiveness, targeting 120 FPS through CPU+GPU acceleration.

Briefing

A new open-source code editor called Zed is making a bid for developers’ attention by pairing extreme speed with built-in AI assistance—without requiring a subscription to a single vendor. Built in Rust by the team behind the Adam editor, Zed is designed to feel “game-like” in responsiveness, targeting low insertion latency, fast startup, and a smaller memory footprint than competing editors. It achieves performance by leveraging all CPU cores and also using the GPU through its own GPU framework, rasterizing the entire window to deliver smooth 120 frames per second.

Beyond raw speed, Zed’s differentiator is AI integration out of the box. It supports GitHub Copilot, but it also lets users bring their own OpenAI API key to chat about code directly inside the editor. That approach shifts costs from a fixed editor subscription to usage-based pricing tied to token consumption. Developers can also select a code block and ask the AI to refactor it, with rewritten code appearing immediately in the workspace. For teams, Zed adds real-time collaboration on the same code base, and the company has already raised $10 million, signaling intent to build a broader product line—possibly including paid features for collaboration.

The broader context is a market racing to combine coding tools with AI. Microsoft’s VS Code remains a dominant editor, while other AI-enabled editors—like Cursor and JetBrains AI—are also competing for mindshare. At the same time, there’s skepticism that AI will replace programmers any time soon; complaints about ChatGPT’s coding performance and the long history of overhyped tech timelines have kept expectations grounded. In that environment, Zed positions itself less as a replacement for developers and more as a faster, AI-augmented workflow.

Still, Zed has clear constraints that limit its immediate reach. The editor is not extensible today, meaning missing features can’t be patched via plugins; the team plans a plugin API in the future, but it’s expected to be Rust-based and likely smaller in ecosystem scale than the JavaScript plugin model that powers VS Code. The biggest practical limitation is platform availability: Zed is currently available only on macOS, with Linux and Windows support planned once the team can justify a viable business model.

For now, Zed looks like a clean, minimal editor with strong performance and a frictionless path to AI-assisted refactoring. Whether it becomes a true VS Code challenger depends on two things: expanding beyond macOS and building an ecosystem—especially extensibility—that developers can rely on as their workflows evolve. If programming remains a human-led activity in the near term, Zed’s combination of speed and AI could make it a serious contender.

Cornell Notes

Zed is a new open-source code editor built in Rust that prioritizes speed and AI-assisted coding. It delivers low latency and high responsiveness by using all CPU cores plus GPU acceleration to rasterize the window and target 120 FPS. The editor integrates AI directly: it supports GitHub Copilot and also allows users to supply their own OpenAI API key for in-editor code conversations and refactoring. Costs can shift from a fixed subscription to token-based usage. The main drawbacks are limited extensibility today and current availability only on macOS, with Linux and Windows support planned later.

What makes Zed unusually fast compared with other editors?

Zed is engineered for responsiveness: it targets low insertion latency, quick startup, and reduced memory use. It achieves this by leveraging all CPU cores and also using the GPU via its own GPU framework. The editor rasterizes the entire window in a way likened to a video game, aiming for smooth 120 frames per second.

How does Zed integrate AI, and how does that affect cost?

Zed includes AI features out of the box. It supports GitHub Copilot, and it can also use a user-provided OpenAI API key to enable chat about code inside the editor. Instead of paying a subscription tied to the editor’s AI, users can pay based on token usage through their own OpenAI account.

What kinds of AI actions can developers perform inside Zed?

Developers can highlight a portion of code and ask the AI to refactor it. The editor rewrites the code directly in the workspace, creating an immediate “before and after” effect. It also supports in-editor conversations about code, not just one-off completions.

What collaboration features does Zed offer?

Zed includes a team-oriented feature for real-time collaboration, allowing multiple people to work on the same code base simultaneously. The company has raised $10 million and may add paid collaboration capabilities later.

Why might Zed struggle to replace VS Code right now?

Two major limitations stand out. First, Zed isn’t extensible today, so missing functionality can’t be filled by plugins; a plugin API is planned but is expected to be Rust-based and likely smaller than the JavaScript ecosystem around VS Code. Second, Zed is currently available only on macOS, with Linux and Windows support planned after the team finds a viable business model.

Review Questions

  1. Which technical choices (CPU/GPU usage and rendering approach) are central to Zed’s performance claims?
  2. How does using a user-supplied OpenAI API key change the AI cost model compared with a fixed subscription?
  3. What two current constraints—one ecosystem-related and one platform-related—most limit Zed’s adoption today?

Key Points

  1. 1

    Zed is an open-source, Rust-built editor designed for very low latency and high responsiveness, targeting 120 FPS through CPU+GPU acceleration.

  2. 2

    Zed integrates AI directly: it supports GitHub Copilot and also enables in-editor code chat and refactoring via a user-provided OpenAI API key.

  3. 3

    AI refactoring works by selecting code and asking for changes, with rewritten code appearing immediately in the editor.

  4. 4

    Real-time collaboration for teams is included, and the company has raised $10 million, suggesting future expansion and possible paid features.

  5. 5

    Zed is not extensible today; planned plugin support may arrive via a Rust-based API that could be smaller than VS Code’s JavaScript plugin ecosystem.

  6. 6

    Zed is currently macOS-only, with Linux and Windows support planned once a viable business model is established.

Highlights

Zed’s performance strategy uses all CPU cores plus GPU acceleration to rasterize the full editor window, aiming for 120 FPS.
Built-in AI isn’t limited to one subscription: users can bring their own OpenAI API key for token-based usage.
The editor can refactor selected code on demand, rewriting code immediately in the workspace.
The biggest adoption blockers are current macOS-only availability and the lack of extensibility until a planned plugin API arrives.

Topics

Mentioned