September 2026

Splunk Agent Observability is an observability, evaluation, and guardrail platform for GenAI and agentic applications.

Agent Observability is now available as a SaaS deployment integrated with Splunk Observability Cloud.

To get started on building safe and reliable AI applications, contact your Splunk team for access and see Splunk Agent Observability (SaaS).

Monitor, control, and alert on the token usage and costs of your AI agent and AI infrastructure providers. See Splunk AI Token and Cost Monitoring.

Create a custom role to enable non-admin users to create, update, or delete business transaction rules and endpoint rules. See:

• Configure business transaction rules

• Create endpoint rules to monitor related requests

Monitor the endpoints related to your application or service with an enhanced detail panel that includes RED metric charts, top 6 traces with errors, and top 6 longest traces. See:

• Monitor endpoints related to your application

• View endpoints related to your service

Monitor your application instances and service instances with an enhanced detail panel that includes RED metric charts, top 6 traces with errors, and top 6 longest traces. See:

• Monitor instances related to your application

• View instances related to your service

Version 1.44.0

• Updated browser execution to Chrome and Chromium 151.
• Updated Runner runtime components to address security issues and maintenance needs.

Version 1.39.0

• Restored NTLM authentication support for HTTP tests.

Observability Logs provides an integrated logs experience in Splunk Observability Cloud. This cloud-only feature uses Splunk Cloud stack for administration and log management.

With Unified Identity, centralized role-based access control (RBAC), and data-level access control, you can view logs alongside metrics and traces and use them in existing Metrics, Events, Logs and Traces (MELT) workflows. Supported onboarding paths include OpenTelemetry and Data Manager. See Observability Logs in Splunk Observability Cloud.

In Splunk Cloud Platform 10.6, you can request and set up a free edition of Splunk Observability Cloud directly from your Splunk Cloud Platform environment.

This release includes the following enhancements:

• Self-service free edition provisioning: If you have the admin or sc_admin role, you can use the Discover Splunk Observability Cloud app to request and provision a Splunk Observability Cloud trial organization.

• Automated log discovery: After you access Splunk Observability Cloud, you are guided through an interactive Log Discovery view that enumerates paired Splunk Cloud indexes and identifies OpenTelemetry (OTel) compliant log sources.

See Get Free Edition from Splunk Cloud Platform.

The definition and SignalFlow logic are now improved to reduce false positive rates and alert noise for the following Autodetect detectors:

• K8s cluster node readiness

• K8s cluster memory utilization is high

• K8s cluster CPU utilization is high

• K8s node memory utilization is high

• K8s node CPU utilization is high

See List of Autodetectors.

This release includes Detector Optimization for Splunk Observability Cloud. Detector Optimization analyzes user-created detectors on a weekly schedule and provides prioritized recommendations to reduce alert noise, identify inactive or muted detectors, and improve detector configuration. You can review expected impact, preview changes, apply or dismiss recommendations, and use bulk actions to disable or delete selected detectors. Recommendations are not applied automatically. See Optimize Detectors.