Last9
Unified observability. Triage Faster.·last9.io
Last9 is an observability platform built for high-cardinality telemetry, providing a developer-first control plane to manage logs, metrics, and traces. It supports OpenTelemetry, Prometheus, and over 100 integrations, offering real-time alerting, data pipelines, real user monitoring, and AI-native features like an MCP server. The platform helps engineering teams debug, optimize, and scale systems while controlling telemetry costs.
What it's for
Features 16
- Agentic developer experience with AI-assisted triage
- Auto-discovery of services, Kubernetes, jobs, hosts, and applications
- Coding agent observability (e.g., GitHub Copilot, Codex monitoring)
- Control plane for telemetry management (route, transform, cost controls)
- Correlated logs and distributed traces
- Dashboard snapshots
- Data pipelines and cost controls before data hits storage
- GPU workload utilization and performance monitoring
- High-cardinality dashboards backed by a Prometheus-compatible query engine
- Log exploration, filtering, and correlated telemetry
- MCP server for AI agent access to telemetry data
- OpenTelemetry and Prometheus native ingestion
- Real user monitoring (RUM) for web and mobile with Core Web Vitals and session replays
- Real-time alerting with deduplication, routing, and on-call integrations
- SLOs and SLIs for service reliability monitoring
- Synthetic monitoring / uptime checks
At a glance
Integrations
Complianceself-reported
Pricing
Pro plan is $3250/month for up to 25 billion events, with additional pay-as-you-go usage pricing beyond that quota. Enterprise pricing is custom/quote-based. The site also states 'Start observing for free. No lock-in.' but no distinct free plan is listed among the two priced offers in the structured pricing data.
Described as pay-as-you-go usage pricing for production apps with growing scale.
25 billion events/month included; 90-day retention for metrics; 14-day retention for logs and traces
Quote-based custom pricing for large apps at internet scale.