Live reliability metrics
We measure every conversion. System failures count against us. User errors don't. See the data.
These indicators update every 60 seconds. The API status reflects whether our conversion service is reachable; the queue indicator reflects current job backlog.
API
Conversion queue
Last update
Rolling success rate, plotted against our 99% target line. Daily buckets for 30 and 90 days, weekly buckets for the 365-day view.
Last 30 days
daily
89.54%
Last 90 days
daily
93.56%
Last 365 days
weekly
91.82%
Rolling 30-day breakdown. Categories with no conversions in the window are hidden.
| Category | Success rate | Status |
|---|---|---|
| Video | 90.91% | Below target |
| Audio | 95.65% | Below target |
| Image | 96.43% | Below target |
| Document | 82.26% | Below target |
We separate two kinds of conversion outcomes. Only one counts against our reliability target.
System failures — count against us
Engine crashes, timeouts, out-of-memory conditions, infrastructure errors, network failures, virus-scanner outages, anything we should have handled better. If you saw one of these, we owe you an explanation.
User-input errors — not counted
Files that exceed your plan's size limit, target formats that no engine can write (e.g. some proprietary ebook formats), corrupted input files, password-protected PDFs, quota-exceeded errors, and validation failures. These are the user's problem to fix and don't reflect on the conversion infrastructure.
Admin-driven smoke tests and internal audits are also excluded so a busy QA day doesn't poison the customer-experience signal.
See how our measured reliability stacks up against the closest competitors.