Sonar announced the general availability of SonarQube Server 2026.5 LTA, enabling enterprises to guide, verify, and continuously resolve issues in agentic code entirely on infrastructure they control, including on-premise, air-gapped, and VPC-restricted environments. Building on the success of Sonar Vortex, SonarQube Remediation Agent, and SonarQube Hunter Agent on SonarQube Cloud, SonarQube Server 2026.5 LTA brings Sonar’s full agentic capabilities to self-managed deployments. Enterprises can set context and constraints for coding agents from the outset, independently verify every change, continuously improve code quality and security, and generate the evidence needed for governance and compliance.
Enterprise software development is changing rapidly as agentic software factories leverage AI coding agents to write higher volumes of code, at very high velocity. GitHub recently reported, for example, that monthly commits grew from 1.4 billion in April to 2.9 billion in August, with 130 million pull requests and 24 million new repos. To realize the value of this increased output, companies must ensure the code is production ready, secure, and maintainable. Developers, executives, and regulators need confidence that the agentic software factories are operating with appropriate controls.
“AI has made code creation abundant, but confidence in that code is scarce,” said Ori Yitzhaki, Chief Product Officer at Sonar. “Enterprises should not have to choose between the speed of agentic development and control over their code, data, and governance. With SonarQube Server 2026.5 LTA, they can apply the same guardrails, verification, and governance to agentic development on the infrastructure they control.”
One platform for enterprise agentic code confidence
Rather than stitching together point solutions, teams can set guardrails, verify code, automate remediation, and surface deeper risks in one integrated workflow. With the 2026.5 LTA release, the full power of the SonarQube platform is available whether you want to operate in the cloud, or have a preference for greater sovereignty by operating on infrastructure in your control.
Sonar’s agentic capabilities enable enterprises to:
- Guide agents as code is being written. Sonar Vortex injects the right project context and constraints before the agent writes its first line, then verifies every change in real time. Agents produce better code with fewer tokens and less rework, reducing token consumption by up to 36%, all before a pull request is ever opened.
- Uncover logic flaws traditional analysis misses. As an added layer of security, SonarQube Hunter Agent employs guided reasoning to hunt for broken access control, business logic, and authentication flaws that typical code analyses tend to miss.
- Resolve technical debt automatically. SonarQube Remediation Agent lets teams continuously improve their codebase, generating verified fixes for prioritized backlog issues. Developers retain control over how and when those fixes are committed, choosing their preferred level of automation.
The release also gives enterprises greater control over AI-scale delivery through centralized LLM governance and MCP access, including state-of-the-art models from Anthropic and OpenAI, with support for open-weight models coming soon.
Beyond governing how agents access and use models, enterprises can verify that agentic output is fit to commit. SonarQube and Gitar, Sonar’s recently acquired AI-native code review solution, help teams identify and fix reliability, security, and maintainability issues while generating the evidence needed for enterprise confidence. SonarQube Advanced Security extends that verification to the software supply chain, helping minimize dependency risk. Reachability and taint analysis, automated compliance reporting, and portfolio-wide architecture management provide additional visibility and control across the application estate.
SonarQube gives enterprises the verification and governance they need to scale agentic software factories with confidence, to successfully operate in the AI economy.
