TestGrid ONE
by TestGrid
AI testing agents plus real-device and browser infrastructure, deployable on-premises or air-gapped
TestGrid ONE is an AI platform for software quality engineering that combines testing agents with real mobile devices and browsers in cloud, on-premises or air-gapped setups. It is for enterprise QA and engineering leaders who want AI test generation and self-healing without rewriting existing Selenium, Playwright or Appium suites.
TestGrid ONE is the platform TestGrid made generally available on November 4, 2026 as the successor to its TestOS product, positioned as an operating system for enterprise quality engineering. It joins two things that are usually bought separately: AI agents that create and maintain tests, and the infrastructure the tests run on. The agent layer includes CoTester, which uses vision-language models to generate test cases from product specifications and keeps human approval gates in the loop, and AgentRx, which heals UI locators when the front end changes, alongside agents for visual regression, accessibility auditing, change-impact assessment, root-cause diagnostics and DevSecOps orchestration; the pricing page counts 27+ agents with no per-agent fees. The infrastructure layer offers real iOS and Android devices and Chrome, Firefox, Safari and Edge browsers, including legacy versions, as hosted cloud, private dedicated hosting, on-premises device labs, hybrid setups or air-gapped installs, with MDM support and Virtual USB. Existing suites in Selenium, Playwright, Appium, Cypress, Espresso, XCUITest, WebdriverIO, Detox, Robot Framework and CodeceptJS run unchanged, and 100+ connectors cover GitHub Actions, GitLab, Jenkins, Azure DevOps, Jira, TestRail and Slack. Coverage extends to APIs, performance, visual, audio and IoT testing and packaged apps such as SAP, Salesforce, ServiceNow and Microsoft Dynamics. Failed runs capture video, screenshots, logs, crash reports and network traffic. TestGrid, led by founder and CEO Harry Rao, was bootstrapped and EBITDA positive when TechCrunch profiled it in 2022, and it competes with BrowserStack, LambdaTest, Sauce Labs, Kobiton and HeadSpin on devices and with mabl, testRigor and Tricentis on AI testing.
A QA or engineering director at a regulated enterprise (bank, insurer, utility, telecom) that must test mobile and web apps on real devices inside its own network.
AI test creation and self-healing on real devices without rewriting existing test code or sending device traffic outside the firewall.
At a Glance
- Category
- Developer Tools
- Pricing
- Subscription, Contact for pricing
- Target Market
- CTOs, QA and Test Engineering Leaders, Enterprise Developers, DevOps Teams
- Deployment
- Hybrid, Self-hosted, Cloud-first
Key Features
- ✓CoTester test generation
Generates manual and automated test cases from product specifications using vision-language models, with human approval gates before tests are adopted.
- ✓AgentRx self-healing
Repairs broken UI locators when the front end changes, reducing the maintenance work that usually stalls automation suites.
- ✓Real-device and browser lab
Physical iOS and Android devices plus Chrome, Firefox, Safari and Edge, including legacy versions, hosted or inside your own network.
- ✓Framework-agnostic execution
Runs existing Selenium, Playwright, Appium, Cypress, Espresso, XCUITest and other suites as written, avoiding a proprietary-format lock-in.
- ✓Air-gapped and hybrid deployment
Private cloud, on-premises, hybrid and air-gapped options with MDM support and Virtual USB serve regulated industries with data-residency rules.
- ✓Failure diagnostics
Each failed run captures session video, screenshots, logs, crash reports and network traffic, so engineers can find the cause without re-running.
Capabilities
Use Cases
- •Modernising a legacy automation suite
Run thousands of existing Selenium and Appium tests unchanged while AgentRx heals locators, so the suite stops decaying with every UI release.
- •In-house device lab for a bank
Stand up real iOS and Android devices behind the firewall to test mobile banking apps, biometrics included, without exposing data to a public cloud.
- •ERP and CRM release testing
Validate SAP, Salesforce or Dynamics updates with AI-generated regression tests before each vendor release reaches business users.
- •CI/CD quality gates
Trigger device and browser runs from GitHub Actions, GitLab, Jenkins or Azure DevOps and push failures with evidence into Jira.
- •Accessibility and visual regression checks
Run agent-driven accessibility audits and visual comparisons on every build to catch layout and compliance regressions early.
Ideal For
Best For
- ✓Regulated enterprises that need an on-premises or air-gapped real-device lab
- ✓QA teams with large existing Selenium, Appium or Playwright suites who want AI maintenance without migrating
- ✓Testing packaged enterprise apps such as SAP, Salesforce, ServiceNow and Microsoft Dynamics
- ✓Mixed-skill QA teams that want scriptless, AI-generated test cases alongside coded tests
- ✓Mobile teams needing biometric, image-injection, audio or geolocation testing on physical devices
Not Ideal For
- ✗Small teams on a tight budget: the entry plan is $249 per seat per month for four devices or browsers
- ✗Buyers who want a self-serve trial; TestGrid's trial is a guided proof of concept set up by a solutions engineer over one to two weeks
- ✗Fully disconnected sites that need AI features: on-premises AI features still require outbound access to TestGrid's hosted AI, and the no-outbound option is scriptless only
Deployment
Market Analysis
Pros
- ✓On-premises and air-gapped real-device labs, which most cloud device farms do not offer
- ✓Works with existing test code, lowering switching cost
- ✓Reviewers praise scriptless test writing and plug-and-play setup
- ✓Published entry pricing
Cons
- ✗Reviewers on SourceForge report it sometimes fails to connect to cloud devices or browsers
- ✗A reviewer said the auto-heal feature was slow to arrive and stayed in beta for a long time, so AgentRx maturity is worth testing in a proof of concept
- ✗Independent review volume is small (nine SourceForge reviews), so there is little evidence at large-enterprise scale
- ✗On-premises AI features still need outbound connectivity to TestGrid's hosted AI
Pricing
Starter
$249 per seat/month
- ✓4 devices/browsers
- ✓5,000 AI tokens
- ✓27+ AI testing agents
- ✓AgentRx self-healing
- ✓TestGrid hosted
- ✓24x5 email support
Growth
Contact for pricing
- ✓Unlimited tokens and executions
- ✓SSO, LDAP, MFA
- ✓Site-to-site VPN
- ✓Device pools and reservations
- ✓Dedicated customer success
Custom Device Lab
Contact for pricing
- ✓On-premises or private cloud
- ✓Unlimited users and execution minutes
- ✓Data residency and audit logs
- ✓MDM and Virtual USB
- ✓24/7 device availability
Starter is $249 per seat per month with four devices or browsers and 5,000 AI tokens; a typical generated test case uses 10 to 20 tokens, while self-healing, bug summaries and root-cause analysis use none. Growth and the on-premises Custom Device Lab are custom-priced with unlimited tokens. The trial is a guided proof of concept, not self-serve.
Security & Compliance
Sources
This page was written from 5 sources, 3 on domains other than testgrid.io.
Stay Ahead of the Curve
Weekly enterprise AI insights for technology leaders. No spam, no vendor pitches—unsubscribe anytime.
SubscribeRelated Products
Ampersand
Integration infrastructure that lets AI agents read and write in customer CRMs and ERPs
OutSystems Agent Experience
Governed AI development on OutSystems for Claude Code, Cursor, Codex and Kiro, via MCP
IBM Bob
IBM's agentic coding IDE and CLI for governed Java, IBM i and mainframe modernization, now self-hostable and air-gapped
HackerRank Chakra
AI interviewer that runs hands-on technical interviews and scores how candidates work with AI