Generate, Run, and Self-Heal E2E Tests
Entirely Inside Your Own Infrastructure
Generate standard Playwright (Web) and Maestro (Android/iOS) test code from natural-language specs, run autonomous browser agents, and automatically repair broken selectors when UI changes. Deploy with Docker Compose on a single laptop, on-prem server, or private cloud VPC.
Currently free to use. Issue a 90-day renewable license key instantly with just your email.
- 100% Data Sovereignty
- Target URLs, login credentials, recorded browser videos, and test results stay strictly inside your own Docker host.
- Web, Mobile & API in One Place
- Manage Playwright web tests, Maestro mobile flows, HTTP API checks, and manual test runs in a unified workspace.
- Bring Your Own AI / Local LLM
- Connect your own OpenAI, Gemini, or Claude API key—or run 100% offline with an internal Ollama / vLLM endpoint.
Watch AXIOM in Action: Full System & Feature Walkthroughs
Explore all 8 walkthrough videos with synchronized English captions and interactive chapter navigation—covering the system overview, AI test generation, OODA autonomous execution, AI selector self-healing, Visual Regression Testing (VRT), cron scheduling, manual testing, and enterprise audit logs.
Get a complete bird's-eye view of AXIOM's four core pillars—AI test creation, autonomous browser execution, AI selector self-healing with visual diffing, and unified manual + scheduled operations—all running inside your own infrastructure.
Automate your entire QA lifecycle from spec authoring and Playwright/Maestro code generation to daily regression runs, selector repair, and compliance reporting without ever sending data to an external SaaS.
Video Chapters (Click to Seek)
Select any step below to jump directly to that timestamp
AI analyzes your target URL and requirements to design test scenarios automatically. Switch seamlessly between the GUI Visual Editor, Gherkin BDD syntax, Playwright TypeScript code, and Maestro mobile YAML flows.
Enable PMs and QA analysts to review scenarios in plain language while engineers work directly with standard Playwright and Maestro code in the same workspace.
Video Chapters (Click to Seek)
Select any step below to jump directly to that timestamp
Combine autonomous 'AI Codegen Mode' (where AI observes, decides, and acts on live browser screens) with deterministic 'Static Mode' that runs recorded Playwright code in seconds—complete with video and audit reports.
Instantly verify why a test passed or where it failed using recorded browser videos and OODA reasoning logs, and export PDF/Excel reports for release sign-off.
Video Chapters (Click to Seek)
Select any step below to jump directly to that timestamp
Solve the #1 pain point of E2E testing: brittle selectors breaking on every UI change. See how AXIOM groups failed runs, links with GitHub Issues, refines scenarios, and self-heals broken Playwright selectors.
Slash test maintenance hours when button labels or DOM structures change and keep your regression suite green.
Video Chapters (Click to Seek)
Select any step below to jump directly to that timestamp
Catch CSS layout breaks and overlapping text that DOM assertions miss by comparing scenario screenshots pixel-by-pixel against baseline images, alongside project-wide pass-rate and coverage analytics.
Quantify both functional stability and visual UI integrity before every production deployment.
Video Chapters (Click to Seek)
Select any step below to jump directly to that timestamp
Embed automated testing into your daily engineering rhythm: organize features into test suites, schedule unattended cron runs, reuse shared login credentials, manage upload test files, and send real-time Slack alerts.
Run unattended nightly full regressions or hourly smoke tests and alert your team on Slack the moment an issue appears.
Video Chapters (Click to Seek)
Select any step below to jump directly to that timestamp
Eliminate scattered spreadsheets by managing human-in-the-loop visual checks and exploratory tests using the same test suite assets—complete with pass-rate metrics, person-hour tracking, and one-click export.
Combine automated E2E runs and manual test records in one project to generate complete release sign-off evidence.
Video Chapters (Click to Seek)
Select any step below to jump directly to that timestamp
Built for strict enterprise security and compliance: full audit log tracking with CSV export, scoped API keys for GitHub Actions and MCP coding agents, organization RBAC, and internal LLM configuration.
Scale test automation under strict governance with internal LLMs, scoped API tokens, and immutable audit trails.
Video Chapters (Click to Seek)
Select any step below to jump directly to that timestamp
* All URLs (shop.example.com), repository names, and user accounts shown in the videos are synthetic demo data and contain zero real personal or corporate information.
Core Capabilities
AXIOM generates standard, human-readable Playwright (TypeScript) and Maestro (YAML) code—zero vendor lock-in. Your test suite remains yours forever.
Generate Test Code from Specs
Turn natural-language requirements or Gherkin scenarios into executable Playwright TypeScript code and Maestro mobile flows. Edit and sync everything directly in the browser.
- Playwright (TypeScript) generation & in-browser editor
- Bidirectional Visual Editor & Gherkin BDD sync
- Unified manual test tracking & Excel/PDF export
AI Self-Healing & Root-Cause Analysis
When a UI update breaks a selector, AXIOM analyzes the live DOM and execution logs to propose and apply selector fixes automatically, or file a GitHub Issue in one click.
- Automated selector & scenario self-healing
- AI failure root-cause summary & OODA logs
- Visual Regression Testing (VRT) pixel diffing
Unified Web, Mobile & Scheduled Runs
Execute Playwright web tests and Maestro mobile flows in parallel. All recorded videos, screenshots, and logs are stored locally inside your server's object storage.
- Parallel Playwright & Maestro test runner
- Cron scheduler with Slack, Webhook & Email alerts
- Full browser video recording & audit reports
Built for Every Environment
The exact same 3-step Docker Compose workflow scales from a single laptop to air-gapped enterprise servers.
Individual Developers & Small Teams
Spin up Docker Compose on your laptop in minutes. Use your own LLM API key or local Ollama—no per-run cloud markup or hidden SaaS fees.
Growing Engineering & QA Teams
Host AXIOM on a single internal VM or cloud instance to share scenarios, test suites, and video evidence across your team with RBAC and Slack/Webhook alerts.
Enterprise & Air-Gapped Networks
Deploy inside private VPCs or on-premises networks without exposing staging environments to the internet. License keys are verified offline via RS256 cryptographic signatures.
Zero-Trust Data Architecture
Every AXIOM container runs inside your private Docker network. Your target URLs, login credentials, screenshots, and test results are never transmitted to our servers.
Inside Your Server (Docker Compose)
Web UI & Go API Server
Next.js management dashboard and Go backend API. Cryptographic license verification runs locally here.
Web UI :3001 / API :8081
Autonomous Test Runner
Accesses your target applications on your internal network to execute Playwright and Maestro tests.
Playwright / Maestro
Local Database & Object Storage
Stores test results, auth profiles, recorded videos, and screenshots. Bound strictly to localhost.
PostgreSQL / Redis / MinIO
External Network Calls
When AI features are used, AXIOM communicates directly with your configured AI provider (OpenAI / Google / Anthropic)—or stays 100% internal when using Ollama. Once per day, a lightweight version check sends anonymous aggregate counts (no URLs, code, or test data), which can be disabled via AXIOM_TELEMETRY_DISABLED=true.
Offline Cryptographic License Verification
License keys are verified locally using an embedded RS256 public key, allowing full operation even in air-gapped environments.
Read Communication Specification →Issue Your Free License Key
Enter your email address to display your license key immediately on this page. Keys are valid for 90 days and can be renewed for free starting 14 days before expiration.
License Key Issued
Submitting the same email up to 14 days before expiration returns your current key; within 14 days of expiration it issues a fresh key. Follow the Quickstart Guide below to install.
Quickstart (3 Steps)
Run on any machine with Docker Engine 24.0+ and Docker Compose v2 (Linux recommended).
1.Download configuration files
curl -fsSL https://axiom-qa.dev/install.sh | sh && cd axiom-selfhostedCreates docker-compose.prod.yml and .env inside ./axiom-selfhosted with auto-generated database/storage secrets.
2.Set your license key in .env
vi .envPaste your issued license key into the AXIOM_LICENSE_KEY= line using any text editor.
3.Start the containers
docker compose -f docker-compose.prod.yml up -dThe first launch may take a few minutes while container images are downloaded.
4.Initial Setup
Open
http://localhost:3001/setupin your browser to create your administrator account. If installed on a remote server, replace localhost with your server's IP or hostname.
Frequently Asked Questions
Is AXIOM really free to use?
Yes. AXIOM is currently free to use. When you use AI features with cloud providers (OpenAI, Gemini, Claude), standard API token usage is billed directly by your chosen provider—or $0 if you use local Ollama.
Can individual developers use it?
Yes. Individual developers, open-source maintainers, and enterprise teams follow the exact same setup. Only an email address is required to issue a license key.
Are my target URLs, login credentials, or test results sent externally?
Never to our servers. All databases, credentials, recorded browser videos, and screenshots remain on the server where you run AXIOM. When AI features are invoked, only the prompt context needed for generation/analysis is sent to your configured AI endpoint.
What telemetry is sent to the operator's server?
Once per day, AXIOM checks for new versions and sends lightweight aggregate counters (instance ID, license ID, version, OS, and total run counts—never test contents or URLs). You can disable this completely by setting AXIOM_TELEMETRY_DISABLED=true in .env.
How long is a license key valid?
Each license key is valid for 90 days. Starting 14 days before expiration, you can issue a new 90-day key for free using the same form on this page.
What happens if my license key expires?
AI generation/self-healing, scheduled runs, and new manual test runs pause until a new key is entered. All existing test scenarios, code, and historical results remain intact and viewable.
Can it run in an air-gapped (offline) network?
Yes. License keys are verified offline inside your AXIOM container via RS256 signatures. Pre-load the Docker images and point the AI configuration to an internal Ollama or vLLM server.
Which AI models are supported?
OpenAI (GPT-4o, etc.), Google Gemini, Anthropic Claude, and local/self-hosted models via Ollama or OpenAI-compatible endpoints.
What are the recommended server specs?
Minimum 4 vCPU and 8 GB RAM. For running multiple parallel browser workers, 16 GB RAM or higher is recommended.