Migrate from BrowserStack, Sauce Labs or LambdaTest
Selenium and Appium tests talk to a remote WebDriver endpoint, so they are not tied to one cloud. Moving a suite from BrowserStack, Sauce Labs or LambdaTest to TestingBot means changing three things:
- the remote URL and credentials,
- the vendor options block in your capabilities, which becomes
tb:options, - a few capability names inside that block.
Your WebDriver and Appium commands stay as they are, and so do most of the JavaScript executor calls you use to mark a test as passed or failed: the TestingBot hub runs them as their TestingBot equivalent.
Moving local tests to the cloud for the first time? Migrate your existing tests covers that case.
Hub URL and credentials
Point your tests at the TestingBot hub. Selenium and Appium use the same endpoint:
https://hub.testingbot.com/wd/hub
TestingBot authenticates with a key and secret, which you find in the member area. Pass them in one of two ways:
- In the URL, the way most cloud setups already do it:
https://KEY:SECRET@hub.testingbot.com/wd/hub. Your client sends them as HTTP Basic authentication. - Inside
tb:options, askeyandsecret.
If your code reads its credentials from environment variables, the quickest change is to point those variables at your TestingBot key and secret. Remove the old username and access key from the capabilities, so they are not sent along with the test.
Rename the vendor options block to tb:options
In W3C capabilities, each cloud reads its own settings from a vendor-prefixed block:
bstack:options, sauce:options or LT:Options.
TestingBot reads tb:options. Rename the block, then rename the keys inside it with the
mapping table below. Settings left in another vendor's block are not applied.
from selenium import webdriver
options = webdriver.ChromeOptions()
options.browser_version = 'latest'
options.platform_name = 'WIN11'
options.set_capability('tb:options', {
'name': 'Checkout flow',
'build': 'nightly-2026-11-16',
'screen-resolution': '1920x1080',
'idletimeout': 120,
})
driver = webdriver.Remote(
command_executor='https://KEY:SECRET@hub.testingbot.com/wd/hub',
options=options
)
ChromeOptions options = new ChromeOptions();
options.setBrowserVersion("latest");
options.setPlatformName("WIN11");
Map<String, Object> tbOptions = new HashMap<>();
tbOptions.put("key", System.getenv("TB_KEY"));
tbOptions.put("secret", System.getenv("TB_SECRET"));
tbOptions.put("name", "Checkout flow");
tbOptions.put("build", "nightly-2026-11-16");
tbOptions.put("screen-resolution", "1920x1080");
tbOptions.put("idletimeout", 120);
options.setCapability("tb:options", tbOptions);
WebDriver driver = new RemoteWebDriver(
new URL("https://hub.testingbot.com/wd/hub"), options
);
const { Builder } = require('selenium-webdriver');
const chrome = require('selenium-webdriver/chrome');
const options = new chrome.Options();
options.setBrowserVersion('latest');
options.setPlatform('WIN11');
options.set('tb:options', {
key: process.env.TB_KEY,
secret: process.env.TB_SECRET,
name: 'Checkout flow',
build: 'nightly-2026-11-16',
'screen-resolution': '1920x1080',
idletimeout: 120
});
(async () => {
const driver = await new Builder()
.usingServer('https://hub.testingbot.com/wd/hub')
.setChromeOptions(options)
.build();
})();
options = Selenium::WebDriver::Chrome::Options.new
options.browser_version = 'latest'
options.platform_name = 'WIN11'
options.add_option('tb:options', {
'key' => ENV['TB_KEY'],
'secret' => ENV['TB_SECRET'],
'name' => 'Checkout flow',
'build' => 'nightly-2026-11-16',
'screen-resolution' => '1920x1080',
'idletimeout' => 120
})
driver = Selenium::WebDriver.for(:remote,
url: 'https://hub.testingbot.com/wd/hub',
options: options
)
ChromeOptions options = new ChromeOptions();
options.BrowserVersion = "latest";
options.PlatformName = "WIN11";
options.AddAdditionalOption("tb:options", new Dictionary<string, object>
{
["key"] = Environment.GetEnvironmentVariable("TB_KEY"),
["secret"] = Environment.GetEnvironmentVariable("TB_SECRET"),
["name"] = "Checkout flow",
["build"] = "nightly-2026-11-16",
["screen-resolution"] = "1920x1080",
["idletimeout"] = 120
});
IWebDriver driver = new RemoteWebDriver(
new Uri("https://hub.testingbot.com/wd/hub"), options
);
The capabilities generator builds this block for any browser and language.
Capability mapping
The middle column lists names that test code written for other clouds commonly uses for each setting. Put the
TestingBot name inside tb:options. Every TestingBot option is described on
Test Configuration Options and, for mobile,
Appium options.
| Setting | Names you may have now | TestingBot (in tb:options) |
|---|---|---|
| Credentials |
userName, username, user, accessKey
|
key and secret, or credentials in the URL (see above) |
| Test name |
sessionName, name
|
name |
| Build |
buildName, build
|
build |
| Project |
projectName, project
|
project |
| Tags | tags |
groups; tags is accepted too |
| Screen resolution |
resolution, screenResolution
|
screen-resolution, for example 1920x1080
|
| Video recording |
video, recordVideo
|
screenrecorder. On by default; set false to turn it off |
| Screenshots | recordScreenshots |
screenshot |
| Idle timeout | idleTimeout |
idletimeout (seconds); idleTimeout is accepted too |
| Maximum test duration | maxDuration |
maxduration (seconds); maxDuration is accepted too |
| Selenium version |
seleniumVersion, selenium_version
|
selenium-version |
| Appium version | appiumVersion |
appiumVersion (same name) |
| Time zone |
timezone, timeZone
|
timeZone |
| Tunnel |
local, localIdentifier, tunnel, tunnelName, tunnelIdentifier
|
tunnel-identifier, with TestingBot Tunnel running |
| Physical device |
realMobile, isRealMobile
|
realDevice: true
|
| Uploaded app |
bs://…, storage:filename=…, lt://…
|
tb://… after you upload the app to TestingBot, or a public URL |
Settings that only exist on another cloud (their own debugging, logging or network capture switches) have no TestingBot counterpart in this table. Remove them. What TestingBot records for every test, starting with a video, is described on test results.
Browser and operating system
The standard W3C capabilities browserName, browserVersion and
platformName keep their names. Only the values can differ:
-
Operating system. Use a TestingBot platform name in
platformName, for exampleWIN11,WIN10,SEQUOIAorSONOMA. The names Selenium 4 sends, such asWindows 11andmacOS 14, are translated. Do not split the OS into a family and a version (anosplusosVersionpair): TestingBot does not read the separate version, so"Windows"with"11"does not pin Windows 11. The full list is on the browsers page. -
Browser version. A number, or
latest,latest-1,latest-2and so on. -
Edge.
MicrosoftEdge,edgeandmicrosoft edgeall select Edge.
Mobile apps with Appium
Appium tests use the same hub URL. Keep the appium: capabilities you have
(appium:deviceName, appium:platformVersion,
appium:automationName), then:
- Upload your app to TestingBot Storage and set
appium:appto thetb://reference it returns. - Set
realDevice: trueintb:optionsto run on a physical device instead of a simulator or emulator. - Check the device names you use against the device selection rules, since each cloud names its devices differently.
Test status commands that keep working
Many suites report their result from inside the test with a JavaScript executor call. For Selenium and Appium sessions, the TestingBot hub recognises the common forms and does the equivalent TestingBot action, so these calls need no changes:
| Call in your code | What TestingBot does |
|---|---|
browserstack_executor: with action setSessionStatus,lambdatest_executor: with action setTestStatus,lambda-status=passed / failed,sauce:job-result=passed / failed
|
Marks the test passed or failed. A reason or remark becomes the status message |
browserstack_executor: with action setSessionName,lambda-name=…, lambdaUpdateName,sauce:job-name=…
|
Renames the test |
lambda-build=…,sauce:job-build=…
|
Files the test under that build |
lambda-test-tags=…,sauce:job-tags=…
|
Adds tags (comma-separated) |
browserstack_executor: with action annotate,lambdatest_executor: with action stepcontext,sauce:context=…
|
Adds a step to the test log |
browserstack_executor: with action getSessionDetails,lambdatest_executor: with action getTestDetails
|
Returns the TestingBot session details as a JSON string: session and test id, name, build, status, browser, platform and a link to the test |
lambda-breakpoint=true,sauce:break
|
Pauses the test so you can take over in the browser, like tb:break
|
The lambdatest_action: and lambda-hook: prefixes are read the same way as
lambdatest_executor:. Any other action with these prefixes is answered with success and does nothing,
so it does not fail your test. A call whose JSON does not parse, or has no action, returns an error.
For new code, use the TestingBot form directly, for example
driver.execute_script('tb:test-result=passed'), or set the result after the test with the
REST API. Both are described on Setting Name and Status.
Playwright
Playwright tests connect to a different endpoint, wss://cloud.testingbot.com/playwright, described on
Playwright testing. That endpoint only understands
testingbot_executor commands, so change the prefix of your status calls and pass the result as a boolean:
await page.evaluate(_ => {}, `testingbot_executor: ${JSON.stringify({
action: 'setSessionStatus',
arguments: { passed: true, reason: 'Checkout completed' }
})}`)
setSessionName works the same way. More examples are on
mark Playwright test status.
Local and staging sites
If your tests reach a site that is not public, replace your current tunnel client with TestingBot Tunnel. There are two ways to send a test through it:
- Point the test at the tunnel's local port,
http://localhost:4445/wd/hub, instead of the hub URL. - Keep the hub URL, start the tunnel with
--tunnel-identifier <id>and set the same value astunnel-identifierintb:options. This is also how you pick one of several tunnels.
A tunnel name or identifier from your previous provider has no meaning on TestingBot; replace it with the identifier you give TestingBot Tunnel.
Migration checklist
- Remote URL set to
https://hub.testingbot.com/wd/hub. - TestingBot key and secret in the URL or in
tb:options; old credentials removed. -
bstack:options,sauce:optionsorLT:Optionsrenamed totb:options, with keys renamed per the mapping. -
platformNameset to a TestingBot platform; no separate OS version. - Apps uploaded to TestingBot and referenced with
tb://. - Playwright status calls switched to
testingbot_executor. - One test run end to end, checked on the dashboard for name, build and status.
If a test fails to start, the test errors page explains the messages the hub returns.
Frequently asked questions
Do I have to rewrite my tests to move them to TestingBot?
No. Your WebDriver and Appium commands stay the same. What changes is the remote URL, the credentials and the vendor options block in your capabilities, which becomes tb:options. See the capability mapping.
Can I keep my bstack:options, sauce:options or LT:Options block?
No. TestingBot reads its settings from tb:options. Rename the block to tb:options and rename the keys inside it with the mapping table; settings left in another vendor's block are not applied.
Do my existing setSessionStatus or sauce:job-result calls still work?
Yes, for Selenium and Appium. The TestingBot hub runs the common BrowserStack, LambdaTest and Sauce Labs JavaScript executor commands as their TestingBot equivalent: test status, name, build, tags and log annotations. See test status commands.
Does this also work for Playwright?
Partly. Playwright connects over a different endpoint, which only understands testingbot_executor commands. Change the prefix of your status calls to testingbot_executor. See Playwright.
How do I test sites behind a firewall after migrating?
Replace your current tunnel client with TestingBot Tunnel. Either point your tests at the tunnel's local port, or start the tunnel with --tunnel-identifier and set the same value as tunnel-identifier in tb:options.