New Relic AI Monitoring

platform

proprietary-SaaS(closed)

license

commercial(Standard/Pro/Enterprise tiers)

ingestion

OpenTelemetry-compatible(open standard, not open product)

source

closed

New Relic AI Monitoring is a commercial module of the proprietary New Relic SaaS platform, with no source published. Support for ingesting OpenTelemetry data does not make the product itself open.

• https://docs.newrelic.com/docs/ai-monitoring/intro-to-ai-monitoring/ recorded 2026-09-29

  Docs intro: "AI monitoring is our solution for application monitoring (APM) for AI." and "Parse information about completion, prompt, and response tokens". Searched the body for self-host, open source and license: none.

• https://newrelic.com/platform/ai-observability recorded 2026-09-29

  Product page for a hosted platform module: "tracking request tokens used and setting custom alerts". Searched the visible copy for self-host and on-prem: none; the open source and license strings in the body are script and navigation boilerplate, not a repository or license offer.

New Relic AI Monitoring is a module of the wider New Relic platform, so adoption reflects the platform's own reported reach - third-party estimates of its user and business base - rather than a count of the module itself. New Relic publishes no customer or usage figure specific to the module.

• https://newrelic.com/platform/ai-monitoring recorded 2026-06-04

  AI Monitoring module of the broad New Relic observability platform; enterprise tiers

• https://newrelic.com/platform/ai-monitoring recorded 2026-08-14

  The URL redirects to newrelic.com/platform/ai-observability and the page is branded AI Observability. It sells the module inside the hosted New Relic platform and publishes NO customer, user or business count anywhere on it - which is what keeps the adoption reading at reported traction rather than a numeric count.

New Relic AI Monitoring covers full-stack AI tracing, including the MCP request lifecycle and tool-call latency, token cost tracking, and quality detection for bias, hallucination and toxicity, with built-in integrations for OpenAI, Pinecone and LangChain. It lacks the prompt-versioning and dataset-evaluation depth of evaluation-first tools, but tracing, cost and quality checks in one view put it level with Langfuse.

• https://docs.newrelic.com/docs/ai-monitoring/intro-to-ai-monitoring/ recorded 2026-09-29

  The AI monitoring introduction says the agent can "Parse information about completion, prompt, and response tokens" and "Correlate negative or positive feedback about a response from your end users", with supported models "from vendors like OpenAI, Bedrock and DeepSeek".

• https://newrelic.com/platform/ai-monitoring recorded 2026-09-29

  The AI Monitoring product page (redirected to the AI Observability page) lists "Use built-in integrations for models like OpenAI, vector data like Pinecone, and frameworks like LangChain", "Take advantage of New Relic AI observability capabilities. No new instrumentation needed.", "the entire Model Context Protocol (MCP) request lifecycle automatically.", "Spot and solve issues like bias, hallucination, and toxicity", and "by tracking request tokens used and setting custom alerts". OpenTelemetry appears only in site navigation there, not as an AI integration.