Semaloop vs Revyl
Revyl is a mobile platform pairing vision-based tests with observability and a map of your app. Semaloop runs natural language tests on real iPhones and Android emulators against every build you ship.
What Revyl is
Revyl runs iOS and Android apps on cloud devices, takes natural language tests with no selectors, and pairs each run with observability — step traces, network, state diffs — plus Atlas, a living map of your app’s screens and flows. It also exposes a Python SDK for programmatic control. Of everything here it sits closest to us.
Where they differ
| Semaloop | Revyl | |
|---|---|---|
| Test authoring | Natural language specs | Natural language tests, no selectors |
| Centre of gravity | Regression on every build | Observability and a map of app behaviour |
| Programmatic control | MCP server and REST API | A Python SDK |
| Deciding pass or fail | A separate judging step against the spec | Run evidence you gate on |
| Non-visual checks | Audio output is in scope | Vision-based |
Choose Revyl if
- You want a behavioural map of the app as much as a pass or fail, and will act on run evidence yourself.
- A Python SDK fits how your team wants to drive things.
- Their device coverage lines up better with your matrix — worth checking directly, since both products move quickly.
Choose Semaloop if
- You want a verdict you can trust without reading the trace, with the trace there when you need it.
- Audio, permissions, push and in-app purchase flows are part of what you need covered.
- You want to model your bill yourself before speaking to anyone.