A focused tool for the people who test software.
Orbis QA is built for SQAs. The goal is a calm, reliable workflow that turns plain language into real browser runs, with the privacy and predictability of software that lives on your machine.
Most QA tooling is built for engineers. It assumes you will write code, maintain selectors, fight flakes, and rewrite fixtures after every release. The people who do QA full time often spend more energy keeping the suite alive than testing the product.
Orbis takes a different position. You describe a flow in plain language. An agent reads the page, decides the next action, takes it, and verifies the result. When something fails, a separate reasoning step decides whether to retry, skip, or stop. Every run produces a self-contained report you can share.
The product runs on your laptop. The scenarios stay on your laptop. The model key sits in your operating system keychain and is used only at the moment of a call. No cloud, no account, no telemetry by default.
Four positions we will not compromise on.
Scenarios, runs, and screenshots live on your machine by default. The product works fully offline.
Your LLM key talks to your provider directly. Orbis never sits in the middle of a request.
Tests are written once in plain language and keep working through UI changes.
Per-run cost caps and a clear cost ledger so the bill never surprises you.
Questions, feedback, or partnership?
Every message is read by the team.