Grafana Labs launches Grafana 13 with observability updates

Grafana Labs has launched Grafana 13 and announced a set of open source observability updates as its user base passes 35 million.

The release focuses on updates to Grafana's dashboarding software, a redesigned Grafana Loki architecture for log workloads, and new work on OpenTelemetry deployment for Linux and Kubernetes. Grafana Labs has also acquired Logline, a startup focused on search across large-scale log data.

The launch comes as engineering teams face growing pressure to manage rising volumes of telemetry data while keeping observability systems interoperable. In its 2026 Observability Survey, Grafana Labs found that more than 77% of organisations rely on open source or open standards for observability, while more than 38% still cite complexity as their main challenge.

Grafana 13 adds suggested dashboards, layout templates based on common methodologies including DORA and USE/RED, and guided learning paths designed to reduce onboarding friction. Dynamic dashboards are now generally available, allowing users to adapt views based on variables and context instead of maintaining multiple static versions.

The release also introduces changes for governance and software development workflows, including two-way Git workflows across GitHub, GitLab, Bitbucket and Git. It also adds a redesigned dashboard schema, a versioned dashboard API, improved secrets handling, dashboard restore, and advisory tools for change management.

The wider ecosystem now includes more than 170 data sources and 120 visualisation panels, supporting users running mixed environments rather than committing to a single monitoring stack.

Logs overhaul

A major part of the announcement centred on Grafana Loki, the company's log aggregation system. Its architecture has been reworked to reflect changing log use patterns as structured logs and OpenTelemetry become more common.

Rather than serving mainly basic search, log systems are increasingly used for analytical and high-cardinality queries. That shift has increased pressure on performance and infrastructure costs, driving changes to ingestion, scheduling and query execution.

The revised Loki design introduces Kafka-backed ingestion, a redesigned query engine and scheduler, and a new query planner that distributes work across partitions and runs queries in parallel. Grafana Labs says this cuts data scanned by up to 20 times and improves aggregated query performance by up to 10 times.

To strengthen log search further, Grafana Labs has acquired Logline. Its technology is designed for precise searches across large log datasets, such as finding a specific user identifier or error code, a common task for operations and troubleshooting teams.

OpenTelemetry push

Grafana Labs also outlined new work tied to OpenTelemetry, the open standard that has become a focal point for vendor-neutral instrumentation. Many organisations are already using OpenTelemetry or moving towards it, but adoption is still slowed by installation complexity, evolving semantic conventions and operational overhead.

In response, its engineers are working with the wider community on integrated OpenTelemetry packages for Linux that can be installed with a single command. The company also highlighted stronger Kubernetes support through the OpenTelemetry Operator.

Another update affects Grafana Alloy, its distribution of the OpenTelemetry Collector. A new OpenTelemetry Engine mode lets teams configure Alloy using standard OpenTelemetry Collector YAML, giving users a more native OpenTelemetry pipeline while retaining integration with Grafana tooling.

"We're seeing a clear shift in how organizations think about observability. It's no longer about choosing a single vendor; it's about building on open foundations," said Anthony Woods, Co-Founder, Grafana Labs. "Open source, open standards, and an open ecosystem give teams the control and flexibility they need in a world that's only getting more complex. The future of observability will be defined by interoperability and community-driven innovation, not closed systems. What we're announcing at GrafanaCON today helps make that open model not just possible, but practical at scale."

Grafana Labs presented the announcements as part of a broader strategy centred on open source software and open standards, with Prometheus and OpenTelemetry at its core. That approach has helped Grafana become a widely used interface for engineers, site reliability teams and developers managing metrics, logs and traces across varied infrastructure.

"GrafanaCON is where the future of observability gets built in the open," said Torkel Ödegaard, Co-Founder, Grafana Labs. "What you see in these announcements is a direct result of that collaboration. Everything we build is shaped by how people actually run these systems at scale. We continue to invest in open source because it's the most effective way to solve hard technical problems together and push the entire ecosystem forward."