Connect every layer of your AI production stack.

Bring models, agent frameworks, cloud infrastructure, vector stores, and team workflows into one operational view.

  • Plug in fast
  • Auto-instrument
  • End-to-end visibility
  • Built for scale
Observability, model providers, frameworks, and runtimes connecting into Metrune

Frameworks

Build and run agents with the frameworks you love.

Learn more
LA

LangChain

Instrument chains and agents with run-level traces that connect prompts, retrievers, tools, models, and final outcomes.

LL

LlamaIndex

Measure retrieval quality, query latency, tool execution, and generation cost across LlamaIndex applications.

Model Providers

Support for leading model providers and platforms.

Learn more
AN

Anthropic

Trace Claude-powered agents, correlate tool use with outcomes, and monitor model cost and performance by release.

AW

AWS Bedrock

Connect Bedrock workloads to Metrune for cross-model observability, production alerts, and environment-aware cost controls.

GO

Google AI

Monitor Gemini model calls and agent workflows with unified traces, release comparisons, and cost attribution.

OP

OpenAI

Send model requests through Metrune to capture traces, latency, token usage, and errors without changing application behavior.

Observability

Collect, correlate, and act on telemetry across your stack.

Learn more
OP

OpenTelemetry

Join AI traces with the infrastructure telemetry your engineering team already collects and operates.

Runtimes

Deploy and monitor where your agents run.

Learn more
KU

Kubernetes

Deploy versioned agent services to Kubernetes and connect runtime health with model-level traces and release history.