Scaling long-running agents (FastRender)
Cursor reports hundreds of concurrent agents building a web browser from scratch, over 1M lines of code in about a week.
Planner/worker/judge hierarchy lets hundreds of agents work on one codebase for days. The write-up finds in-between structure works best and that prompts matter more than harness details; GPT-5.2 did best on sustained runs.
- Date
- Wednesday 14 January 2026
- Lab
- Anysphere (Cursor)
- Kind
- paper
- Access
- research preview
Figures
| Measure | Value | Measured by |
|---|---|---|
| FastRender browser | >1M lines, ~1,000 files Built in about a week, according to Cursor. Simon Willison's post says the browser is written in Rust. | company |
This is a research experiment, not a shipping product, and the account is Cursor's own. Simon Willison wrote about FastRender in a 2026-01-23 post.
Sources
This record was checked against its sources on 6 October 2026. How we check