Feature benchmarks
GET / — dispatches a request through the router and returns a rendered template — no database access. The `/echo` band runs the same path with a constant handler instead, so what the template render costs can be read off directly rather than inferred.
GET /items — loads one page of the 1,000 seeded rows through each framework's ORM / Active Record layer: 20 items plus a COUNT for the pagination total. The `-fast` band repeats that page with hydration switched off, so the cost of building one object per row can be read directly instead of inferred.
GET /api/items — serves the same page of items as a JSON response rather than HTML, which adds serialization to the ORM work. The `-fast` band repeats that response with hydration switched off, so serialization cost and hydration cost can be told apart.
Latency by endpoint (ms, trimmed mean — lower is better)
| Request | Workload | Azera | Laravel | Symfony | Spiral | CodeIgniter | CakePHP |
|---|---|---|---|---|---|---|---|
GET / | no DB — routing + template only | 0.625 | 3.62 | 1.85 | 6.71 | 1.88 | 1.44 |
GET /echo | no DB — routing + dispatch only | 0.518 | 3.45 | 1.60 | 6.54 | 1.72 | 1.30 |
GET /items | 20 of 1000 items (page 1, + COUNT) | 1.34 | 4.70 | 3.40 | 7.63 | 2.70 | 2.79 |
GET /items-fast | 20 of 1000 items (page 1, + COUNT), hydrated rows not built | 1.25 | 4.42 | 3.31 | 7.49 | 2.58 | 2.70 |
GET /items/1 | 1 item by id | 1.23 | 4.41 | 2.74 | 7.48 | 2.51 | 2.56 |
POST /items | 1 row upserted (sentinel #999999) | 1.41 | 4.45 | 3.68 | 7.52 | 2.69 | 2.72 |
GET /items-qb | 20 of 1000 items (page 1, + COUNT) | 1.17 | 4.24 | 2.42 | 7.12 | 2.55 | 2.19 |
GET /items-qb/1 | 1 item by id | 1.13 | 4.11 | 2.36 | 7.08 | 2.46 | 2.11 |
POST /items-qb | 1 row upserted (sentinel #999997) | 1.24 | 4.24 | 3.23 | 7.24 | 2.70 | 2.22 |
GET /api/items | 20 of 1000 items as JSON | 1.21 | 4.55 | 2.73 | 6.78 | 2.51 | 2.46 |
GET /api/items-fast | 20 of 1000 items as JSON, rows not hydrated | 1.13 | 4.14 | 2.67 | 6.63 | 2.42 | 2.41 |
GET /api/items/1 | 1 item by id as JSON | 1.14 | 4.36 | 2.50 | 6.68 | 2.42 | 2.43 |
POST /api/items | 1 row upserted (sentinel #999998) | 1.22 | 4.27 | 3.43 | 6.76 | 2.51 | 2.59 |
GET /features/aop | no DB — interceptor pipeline | 2.32 | 4.97 | 3.08 | 8.23 | — | — |
GET /features/cache | COUNT(*) of 1000 rows, cached 10s (miss = query) | 1.41 | 4.29 | 2.70 | 7.32 | 2.36 | 2.32 |
GET /features/log | no DB — buffered log handlers | 1.00 | 3.50 | 1.74 | 6.69 | — | — |
GET /features/retry | no DB — retry policy | 1.01 | 3.51 | 1.74 | 6.74 | — | — |
GET /features/pipeline | no DB — middleware pipeline | 0.576 | 3.50 | 1.75 | 6.72 | — | — |
GET /features/db-events | 1 event row INSERTed per request | 1.51 | 4.30 | 2.86 | 7.57 | 2.62 | 2.54 |
GET /features/events | no DB — in-process listeners | 1.48 | 4.01 | 2.19 | 7.28 | 2.56 | 1.87 |
GET /features/validation | no DB — validator run | 0.585 | 4.41 | 2.10 | 6.83 | 2.17 | 1.73 |
GET /features/config | no DB — config lookup | 0.531 | 3.49 | 1.76 | 6.72 | 1.82 | 1.30 |
GET /features/request-scoped | no DB — scoped service resolve | 0.532 | 3.49 | 1.74 | 6.78 | 1.80 | 1.33 |
GET /features/rate-limit | no DB — cache-backed limiter | 0.536 | 3.64 | 1.78 | 6.84 | 1.81 | 1.43 |
Server floor — real nginx + PHP-FPM with pm.max_requests=0: the pool never recycles its worker, so no process is spawned per request — what remains is the FastCGI handshake plus a minimal script. A hello-world endpoint that boots nothing but PHP (floor-php) and a static file through nginx (floor-http) measure exactly that cost — the floor every row stands on. Subtracting that floor leaves the framework's own per-request boot, which FPM still pays for every request even though its worker survives.
Full comparison
This page is a summary. The complete report, with every feature chart and the endpoint table for all six frameworks, is published at: