Prometheus v3.13.0 is a Long Term Support (LTS) release, focusing on stability, security, and performance.
Key Changes
Security & Stability: This version addresses critical security vulnerabilities, including an XSS fix in the UI (CVE-2026-44990) and a crucial change preventing credential forwarding on redirects to different hosts (CVE-2025-4673, CVE-2023-45289). Rule group pagination tokens now use SHA-256 for enhanced security. Several bugfixes improve stability, such as resolving race conditions and correcting PromQL range query evaluations that previously miscounted samples.
Performance Enhancements: Users will notice substantial performance gains. Label evaluation with case-insensitive regex is up to 2x faster. TSDB operations see a 12-15% speedup in queries due to reduced per-sample overhead, and V2 histogram WAL decoding is more efficient, cutting allocations by up to 50%.
New Features & Usability:
An experimental API introduces search endpoints for metric names, label names, and values, greatly simplifying data exploration. PromQL gains experimental min_of() and max_of() scalar functions, along with support for smoothed/anchored rates with native histograms. Query statistics now expose samplesRead for better performance analysis. Scrape configurations offer more flexibility with new internal labels for per-target histogram overrides. UI autocompletion is also improved.
For a detailed list, refer to the official Prometheus release cycle documentation.
Impact for QA Teams
QA teams benefit from enhanced security in monitoring infrastructure, reducing potential vulnerabilities during testing. The significant performance improvements mean faster feedback loops for performance tests and more efficient analysis of monitoring data. New API search endpoints and improved PromQL features streamline debugging and validation of metrics, making it easier to verify system behavior and data integrity. For more on integrating Prometheus, see our article on Grafana Prometheus Monitoring.
