Computed from our production database on an activated-accounts basis: an account counts from its first real API request, not from signing up. We publish rates rather than raw counts.
| Week starting | User-base growth |
|---|---|
| 2026-09-14 (current, partial) | +2.5% |
| 2026-09-07 | +5% |
| 2026-08-31 | +9.3% |
| 2026-08-24 | +8.2% |
| 2026-08-17 | +16% |
| 2026-08-10 | +45% |
| 2026-08-03 | +61.8% |
| 2026-07-27 | +35.1% |
| 2026-07-20 | +24.1% |
| 2026-07-13 | +12.7% |
| 2026-07-06 | +21% |
| 2026-06-29 | +27.7% |
| 2026-06-22 | +16.7% |
An account is activated the first time its API key makes an authenticated request to our gateway. Weekly growth = newly activated accounts during the week ÷ activated base at the start of that week (Monday, UTC). Our homepage shows the average over the full weeks listed here — the rolling 7-day figure above uses the same definition over the trailing seven days, but a single week is noisy in both directions. We count activations rather than registrations because registrations include bots and accounts that never use the product; activation can lag signup, and our per-key activation marker only exists since 2026-08-14 (earlier activations are backfilled from request logs), so mid-August weeks read somewhat hotter than reality. Values are read from the production database and cached for one hour.