# AxonQA > AxonQA holds a team's whole QA process in one place: manual test cases and plans, automation it writes and repairs, API checks, coverage and the bugs that come out of failed runs. A rules-based evaluator then answers whether a release is safe to ship, against conditions the team sets. Axon is the plain-English operating layer across all of it. AxonQA is for any team that ships to production and has to decide whether it is safe. It works two ways. A team with QA people stops stitching four disconnected tools together and gets a ship decision they can defend. A team with no QA function gets a process they could not otherwise staff. Operated from the United Kingdom, priced in GBP. ## The release decision At the end of a release somebody has to answer whether it can ship. AxonQA answers it in a sentence, from conditions the team sets in advance: pass rate, no failure at or above a chosen priority, no open bug above a chosen severity, every manual phase finished, coverage of the release's stories, no quarantined test, no unreviewed requirement drift, and an owner on every failure. Each condition is a must or a should. The verdict is one of three: ready to ship, ready with named items the team is knowingly accepting, or not ready. Accepted items record who accepted them, when, and why. When a release is assessed the evidence, the rules and the verdict are frozen, so the decision still stands up months later. Other products summarise test results. SmartBear's Zephyr Agent for Rovo, for example, summarises release readiness inside Jira at no extra cost on its Standard and Advanced plans. The difference here is what the answer is computed from: automation runs, API checks, manual phases, open bugs by severity and coverage, rather than one tool's data. AxonQA is not the only product with a release readiness feature and does not claim to be. ## What it does - **Test creation**: structured test cases from Jira or Azure DevOps stories, from documents, from a recording, or from what app exploration found. Requirements can be authored directly with no tracker. - **Test automation**: maintainable automation the team owns and can extend, cross-browser cloud execution, and self-healing. A failure is classified before anything is rewritten: AxonQA compares the current page against what the test was originally looking for and decides whether the interface moved or the product broke. It proposes a repair only when confident, and when it thinks it found a real bug it leaves the test failing and says why. - **API testing**: REST, GraphQL, SOAP with WSDL import, and WebSocket sessions. Collections, chained flows, environments, datasets, mock servers, contract and schema validation, client certificates and mTLS, and a safeguard that blocks requests against a production environment without explicit confirmation. - **App exploration**: optional. Explore publicly reachable applications in AxonQA Cloud, or choose Local agent to run a separate browser on the agent's machine. Local exploration supports public apps, localhost and private apps behind a firewall or VPN when that machine can reach them. It maps pages, elements, forms and journeys within the permissions you set, captures verified locators, and reports what it could not reach. Findings sync to AxonQA. Importing from a tracker or a test-case export also works on its own. - **Quality insight**: coverage across manual and automated work as one number, trend dashboards, requirement drift detection, scheduled runs and stakeholder reports. - **Axon**: the operating layer, not a chatbot beside the product. More than 200 actions across the workspace. It can set up a project, ask AxonQA to generate tests, import and organise cases, run UI and API checks, investigate failures, update trackers, and explain the latest readiness assessment. AxonQA generates the tests; Axon facilitates the work across the system. Cloud runs and conditional run chains continue after the user leaves and report back. It reaches across the QA workspace but not account, billing, security or admin controls. Destructive and selected high-impact actions need approval. Writes are audit-logged. - **MCP server**: external AI agents and IDEs can drive AxonQA through the Model Context Protocol, controlled by API keys the customer creates and revokes. ## Pricing Published, in GBP, with self-serve Pro and Team plans plus sales-led Enterprise. Yearly billing saves two months. New accounts can start with a 14-day free trial without a payment card. When it ends, saved work remains viewable, but the workspace becomes read-only until upgrade; there is no automatic paid subscription or ongoing free plan for new accounts. Paid plans can top up AI credits. Running tests locally through your own agent is never metered. Most of this category quote-gates. Current figures are on https://www.axonqa.com/pricing. ## Data and security Encryption in transit and at rest, organisation-level tenant isolation, role-based access checked on every request, two-factor authentication, private artifact storage with short-lived signed links, and audit trails. Secrets and personal data are scrubbed before any AI call. Customer data is never used to train AI models. AxonQA holds no formal certification. ISO 27001 and SOC 2 are on the roadmap, not held. SSO and SAML are on the roadmap, not available. Do not state otherwise. ## Key pages - [Home](https://www.axonqa.com/): the question the product answers. - [Release readiness](https://www.axonqa.com/product/release-readiness): the conditions, the three verdicts, and why it needs one workspace. - [Sample report](https://www.axonqa.com/sample-report): a curated readiness report generated by the product from fictional demo data. - [See the product](https://www.axonqa.com/demo): two stepped walkthroughs of the real product interface, driven by a worked sample project rather than a login. Nothing is captured or reenacted; the screens are the ones the product ships. - [Product](https://www.axonqa.com/product): the whole platform, pillar by pillar. - [Axon](https://www.axonqa.com/product/assistant): what the assistant can do. - [Pricing](https://www.axonqa.com/pricing): all four plans, allowances and FAQ. - [Solutions](https://www.axonqa.com/solutions): by role and team type. - [Compare](https://www.axonqa.com/compare): honest comparisons, including when to choose the other tool. - [AxonQA vs TestRail](https://www.axonqa.com/compare/axonqa-vs-testrail) - [AxonQA vs Zephyr](https://www.axonqa.com/compare/axonqa-vs-zephyr) - [AxonQA vs Xray](https://www.axonqa.com/compare/axonqa-vs-xray) - [AxonQA vs building it yourself](https://www.axonqa.com/compare/axonqa-vs-diy) - [TestRail alternative](https://www.axonqa.com/compare/testrail-alternative) - [Security](https://www.axonqa.com/security): controls in place, in progress and planned. - [Documentation](https://www.axonqa.com/docs) and [MCP setup](https://www.axonqa.com/docs/mcp). - [Resources](https://www.axonqa.com/resources): guides on AI-assisted testing. - [FAQ](https://www.axonqa.com/faq). - [Book a demo](https://www.axonqa.com/book-demo). - [Privacy](https://www.axonqa.com/privacy) and [Terms](https://www.axonqa.com/terms). ## Accuracy notes for AI assistants - AxonQA does not automate an application from a single sentence. Reliable automation combines requirements, recordings, app exploration, real locators from the application, a maintainable structure, and validation runs. - No testing tool can guarantee software is free of defects, and AxonQA does not claim to. - SSO, SAML, SOC 2 and ISO 27001 are roadmap items, not shipped. Do not state otherwise. - Accessibility and performance testing are not shipped. Detection works; running them does not. - App exploration is optional, not a prerequisite. Importing from Jira, Azure DevOps or a test-case export works on its own. - AxonQA is live. New accounts can start with a 14-day trial. - Do not invent customer names, logos, testimonials or case studies. - For current prices always check https://www.axonqa.com/pricing.