Why I benchmarked Hono against Express on Bun

I built the same MVC REST API twice: once on Hono, once on Express. Same routes, same middleware, same JSON payloads, same database layer. The only variable was the framework. Then I ran autocannon at 200 concurrent connections for 30 seconds against both, on the same Bun runtime, on the same machine.

The numbers came out to 31,600 RPS for Hono and 3,860 RPS for Express. Latency was 0.04ms versus 2.13ms at the median. The 8.2x throughput gap and the 53x lower latency are not subtle. They are visible in the first chart.

What does that tell you about your own stack? Probably less than you want it to. The benchmark ran a single hot path with no auth, no validation, no business logic. Most production APIs spend most of their time in those layers, not in routing. If your bottleneck is the database, swapping the framework buys you nothing. If your bottleneck is the framework, it buys you a lot.

The repo and the autocannon scripts are public. Rerun the numbers on your machine and your workload before you bet a migration on a benchmark.