ORBIT

Features

AI-powered testing. Built for your infrastructure.

Create tests, run them across web, mobile and APIs, and investigate every failure in one platform that deploys inside your own environment.

Book a demo Explore features
orbit.internal / run 4821
  • Open account dashboard1.2s
  • Transfer £250 to savings2.8s
  • Check ledger entry created0.9s

Execution log

GET /api/accounts 200 118ms
POST /api/transfers 201 402ms
SELECT ledger WHERE ref=... 0 rows
assert: entry exists -> failed
Mockup. Sample data.

Everything needed to author a test

AI test generation

Generate tests from requirements and feature specs. Add screenshots, designs and product knowledge for more accurate coverage.

Plain-English test steps

Describe actions and expected outcomes in ordinary language using AI actions and checks.

Autonomous discovery

Beta

Give a URL and a testing goal and let Orbit explore, or guide it at decision points.

Visual test editor

Edit, reorder, duplicate and skip steps, or mark them optional, without touching code.

Web, mobile and API workflows

Author web, Android, iOS and API tests side by side in one workspace.

Reusable modules

Share common step sequences across test cases so a change lands everywhere at once.

How it works

Computer-use AI, purpose-built for testing.

Computer-use automation lets Orbit interact with an application the way a person does: reading the interface, clicking, typing and navigating, rather than depending on fixed selectors.

Those actions are driven by a small language model built specifically for software testing. Because the model is narrow in scope, it is optimized for cost per test flow rather than general-purpose reasoning.

  1. 01Testing intent
  2. 02Specialized testing model
  3. 03Computer-use actions
  4. 04Execution evidence

Visual logs

See exactly where a run went wrong

Every run records what happened on screen and underneath it, so a failure traces straight back to the step that caused it.

  1. 00:00Launch buildpass
  2. 00:04Sign inpass
  3. 00:11Submit transferpass
  4. 00:14Verify ledger rowfail

Failure detail

Step 4 failed: expected one ledger row for reference TR-4821, found none. Screenshot, console output and the network call for this step are attached to the run.

ScreenshotConsoleNetwork
Mockup. Sample data.

Video playback

Watch the run as it happened, scrubbed to any step.

Step screenshots

A captured screen before and after each action.

Console logs

Browser and app console output collected per step.

Network activity

Requests, responses and status codes during the run.

Execution timings

Duration for each step and for the run as a whole.

Error details

The failing step, its context and the raised error.

Quality reporting

Track execution history, pass rates, recurring failures, step performance and run cost across suites and over time.

Flexible testing workflows

API and database checks

  • HTTP requests with headers, bodies and response assertions
  • SQL queries against your test databases
  • Endpoint selection from an OpenAPI spec

Data-driven testing

  • CSV imports mapped to test steps
  • Reusable datasets across suites
  • Per-row pass and fail reporting

Variables and shared files

  • Environment constants and secrets
  • Values extracted mid-run and reused later
  • Shared upload files for form and import flows

Advanced logic

  • Conditions and loops inside a test
  • Assertions on state, content and responses
  • Performance thresholds on step timings

Visual checks

Beta
  • Screenshot baselines per step
  • Checks for visual match, layout and images
  • Spelling and language checks on rendered text

Download validation

Beta
  • Validate downloaded file names
  • Check file sizes
  • Inspect file contents after download

Work with Orbit the way your team works

Chat

Ask Orbit to build, change or run a test in a conversation, then review the steps it produced.

Voice

Speak instructions when your hands are busy with the application under test.

Accessibility

Include accessibility considerations in your testing workflows alongside functional checks.

Deployment

Runs inside your infrastructure

Orbit deploys into your own environment, so you control where tests run and how they reach your systems. External models and integrations you choose to connect remain external.

Test organization

  • Folders and tags
  • Priorities per test
  • Suites for grouped runs

Mobile builds

  • Upload APK, IPA and AAB files
  • Pick a build per run
  • Keep build history per suite

REST API

  • Trigger suites from CI/CD
  • Poll run status and results
  • Pass variables per run

Integrations

  • GitHub
  • Jira
  • LambdaTest
  • BrowserStack
  • AWS S3
  • Google Cloud Storage
  • PostgreSQL
  • MySQL
  • MSSQL

Coming soon

Slack, Trello, Asana

See Orbit test your application.

Walk through your own testing workflows with our team and see how they run in Orbit.

Book a demo