March 2026 Summaries

Zeta2 is the updated edit prediction model for Zed, offering a 30% higher acceptance rate than its predecessor, Zeta1. While maintaining its core architecture, Zeta2 introduces significant changes in context building, model training, and evaluation, enhancing its performance and user feedback integration. The model is trained on an extensive dataset of nearly 100,000 examples sourced from open-source repositories, improving its accuracy and latency. The adoption of LSP-based context retrieval allows Zeta2 to make more informed predictions by accessing surrounding code definitions. Users can explore, run, and fine-tune Zeta2's open-weight model available on Hugging Face. Zed also supports multiple edit prediction providers like Mercury Coder and Copilot Next-Edit, catering to different language and editing styles. As the development team focuses on continuous improvement, they are experimenting with Direct Preference Optimization and new prompt formats to enhance model efficiency. Users are encouraged to contribute to future advancements by enabling training data collection, and those interested in furthering the project's goals are invited to join the team.

Zed's engineering team is defined by its wide-ranging codebase and the unique approach to software development, emphasizing full-stack ownership, real-time collaboration, and vertical integration. The company hires engineers through diverse pathways, including traditional applications and open-source contributions, seeking individuals who demonstrate a strong ability to tackle complex problems collaboratively. New hires are quickly immersed in the work, with an emphasis on active participation and real-time problem-solving, often through pair programming. Zed's vertically integrated system means engineers must consider the entire system's cohesion, with a focus on reducing overall complexity rather than specializing in one layer. The company's culture values the balance between rapid shipping and craftsmanship, fostering an environment where engineers are expected to navigate challenges regularly. The hiring process includes technical discussions and pairing sessions that reflect Zed's practical work environment, encouraging potential hires to engage deeply and consistently with the team.

Zed has introduced the Zed Student plan, granting university students free access to Zed's Pro features for one year, to enhance collaborative coding experiences. Zed is designed to facilitate real-time collaboration directly within the code editor, addressing the limitations of traditional tools like screen sharing and pasted snippets. The platform, which is fast, intentional, and open-source, allows students to contribute by filing or addressing tickets, enhancing their skills in languages like Rust. Many of Zed's team members are students or recent graduates, reflecting the company's commitment to integrating students into its development process. The Student plan offers $10/month in AI token credits, unlimited edit predictions, and real-time collaboration, with additional free resources available from providers like Gemini and GitHub Copilot. Alongside this, Zed has launched a Campus Ambassadors program to engage passionate student leaders keen to promote and develop the platform on university campuses. Students worldwide can apply for the plan at zed.dev/education, with seamless transitions for those on free trials or existing Pro subscriptions.

The blog post highlights various lesser-known features of the Zed code editor that enhance productivity and streamline workflows. It introduces methods to open multiple project search tabs and efficiently gather selections for multi-cursor editing. Users can quickly toggle between different settings profiles, allowing for instant configuration changes such as switching themes or adjusting layouts for specific tasks like social media presentations. Editing shortcuts are provided for quickly inserting new lines and deleting characters, while the diff tool aids in identifying differences between code blocks for refactoring. Additionally, the keybinding system enables sequencing multiple actions under a single command, exemplified by automating a distraction-free writing mode. The post encourages readers to share their own tips to further enrich the Zed community's knowledge base.