Status
API status
A live check from your browser, the targets we hold ourselves to, and where to find incident updates.
Checking…
GET https://api.serpkite.com/healthz
- Round trip (last)
- –
- Round trip (median)
- –
- Checks this visit
- 0
- Last checked
- –
Checks from your browser, every minute while this tab is open
Round trip includes your network. Server-side latency for real searches is in the X-Latency-Ms header of every response.
Live
Per-endpoint metrics
Server-side success rate and p50/p95 latency over the last hour, from GET /status. Failed requests are never billed.
Loading live metrics…
Targets
Service level objectives
These are targets from our product spec, not measured results. Measured p50/p95 per endpoint will be published here once the public benchmark runs.
≤ 1.2 s
p50 latency, real-time /search (US/EU)
≤ 2.5 s
p95 latency, real-time /search (US/EU)
99.9%
API uptime
≥ 99%
Success rate (non-empty, correctly parsed)
| Endpoint | p50 target | p95 target | Success | Uptime |
|---|---|---|---|---|
Google Search
/v1/search
|
≤ 1.2 s | ≤ 2.5 s | ≥ 99.0% | 99.9% |
Google News
/v1/news
|
≤ 1.2 s | ≤ 2.5 s | ≥ 99.0% | 99.9% |
Google Images
/v1/images
|
≤ 1.2 s | ≤ 2.5 s | ≥ 99.0% | 99.9% |
Google Videos
/v1/videos
|
≤ 1.2 s | ≤ 2.5 s | ≥ 99.0% | 99.9% |
Google Maps
/v1/maps
|
≤ 1.2 s | ≤ 2.5 s | ≥ 99.0% | 99.9% |
Google Places
/v1/places
|
≤ 1.2 s | ≤ 2.5 s | ≥ 99.0% | 99.9% |
Google Reviews
/v1/reviews
|
≤ 1.2 s | ≤ 2.5 s | ≥ 99.0% | 99.9% |
Google Shopping
/v1/shopping
|
≤ 1.2 s | ≤ 2.5 s | ≥ 99.0% | 99.9% |
Google Scholar
/v1/scholar
|
≤ 1.2 s | ≤ 2.5 s | ≥ 99.0% | 99.9% |
Google Patents
/v1/patents
|
≤ 1.2 s | ≤ 2.5 s | ≥ 99.0% | 99.9% |
Google Autocomplete
/v1/autocomplete
|
≤ 1.2 s | ≤ 2.5 s | ≥ 99.0% | 99.9% |
Google Lens
/v1/lens
|
After benchmark | After benchmark | ≥ 99.0% | 99.9% |
Webpage to Markdown
/v1/webpage
|
After benchmark | After benchmark | ≥ 99.0% | 99.9% |
Custom Search (CSE-compatible)
/customsearch/v1
|
≤ 1.2 s | ≤ 2.5 s | ≥ 99.0% | 99.9% |
Lens and Webpage do more work per call; their latency targets follow the first benchmark run. Parse drift is caught by scheduled canary queries on every vertical, with a 24-hour fix target. Methodology: benchmarks.
History
Incident history
No incidents reported.
Incidents and maintenance windows are posted on status.serpkite.com as they happen, with a write-up afterwards.
Get notified
Subscribe on the status page for email updates on incidents and maintenance. Account-level alerts (low balance, usage thresholds) are in your dashboard settings.
FAQ