Robert Ta describes how Cursor built an agent swarm that reimplemented SQLite in Rust using frontier models as planners and cheaper models as workers. Separating planning from implementation cut costs to 7.9 times lower than using only expensive models, while maintaining near-identical quality.
Cursor had an agent swarm implement SQLite in Rust from scratch against an 835-page manual graded by SQL Logic Test. You could see Fable 5 here with the highest cost, $20,000. Lowest cost, Opus 4.8 plus Composer 2.5, where they split up the workloads to rebuild SQLite. The architecture was planner agents being the frontier models. They decompose goals. Worker agents, fast, cheap models, execute on them. And Wilson Lin says a planner doesn't implement, so its context never fills with low-level detail. Worker never plans, so it can spend all its context on one narrow piece of work. And the Frontier as planner plus cheap worker mix delivered near identical quality at 7.9x lower cost than Frontier only. So the thesis is really that Frontier for planning and orchestration, and then lower level models that cost less for the actual implementation.
