New Relic AI Observability

What it does

New Relic AI Observability takes in telemetry from AI-powered applications, including LLM calls, agent and tool calls, MCP server requests, model predictions, and the surrounding APM, infrastructure, and database signals. It produces dashboards, traces, alerts, and cost views that show response time, quality signals such as prompts and responses, token usage, and where latency or errors occur across the stack.

People use it to troubleshoot slow or failing AI features, follow multi-agent call flows through distributed tracing, and compare models on cost, performance, and quality in one view. Its distinct claim is one platform covering the AI layer alongside apps and infrastructure, with out-of-the-box connections to services such as OpenAI, Pinecone, and LangChain.

It suits engineering and reliability teams running AI applications in production. Starting is free with a perpetual free tier of 100 GB monthly data ingest and one full platform user, with no credit card required. Paid usage applies beyond that, and the platform's own pages note that cost varies if add-on compute features are enabled. Limitations include that most paid plan prices require a sales or pricing request.

What it can do

• Monitor LLM response time, errors, token usage, and cost across models
• Trace agent-to-tool and agent-to-agent call flows with distributed tracing
• Automatically capture the full Model Context Protocol (MCP) request lifecycle
• Compare models on cost, performance, and quality in one view
• Track ML model predictions and drift with out-of-the-box dashboards
• Set custom alerts on token usage and related AI metrics

Works with

• OpenAI
• Pinecone
• LangChain
• OpenTelemetry

Before you choose

• Paid pricing for Standard and Pro editions requires a pricing request; the page does not publish those amounts.
• Advanced Compute add-on features can incur extra charges beyond the base price.
• Data ingest stops and platform access is lost if the free 100 GB monthly limit is exceeded until upgrade or the next month.

Listed under

• Coding
• DevOps and observability

Get started

Free tier includes 100 GB monthly ingest and one full platform user; no credit card required.

1. Create a free New Relic account at the signup page.
2. Connect your AI application's telemetry, using the built-in integrations for supported models and frameworks.
3. Open the AI observability views to confirm LLM, agent, and MCP data is arriving, then set alerts on token usage or latency.

The free tier includes 100 GB of monthly data ingest and one full platform user. Paid charges depend on the selected data, user and compute pricing model. Confirm edition entitlements, retention and usage rates for the intended deployment.

Source: newrelic.com/pricing, checked 30 Oct 2026. Prices can change; check before you buy.

Sources for this overview

Product pages, documentation, and pricing checked 30 Oct 2026. Each source supports part of the overview; availability and prices can change.

• AI Observability for Modern Applications | New Relic ↗newrelic.com
• Transparent Pricing - Start for Free | New Relic ↗newrelic.com

Company website

newrelic.com: product, pricing and policy pages

Web search

Confirms the official site

Social accounts

Accounts the company links, and its posts on X

Human review

Checked by a person

knowledge_base

manual_review

Last checked 30 Oct 2026. Every field keeps its source and date; unknown means we couldn’t confirm it.