How to Cancel Datadog in 2026 (The $23/Host/Month Money Pit)

Updated February 2026 • 6 min read

Datadog charges $23/host/month for Infrastructure, $31/host/month for APM, plus separate charges for logs ($0.10/GB), custom metrics ($0.05 each), and more. A 50-server startup easily hits $5,000-15,000/month. Here's how to cancel and what open-source tools replace it.

🔴 The surprise bill: Datadog is infamous for unexpected bills. Auto-scaling infrastructure, container orchestration, and custom metrics create charges that grow unpredictably. Check your bill before canceling — you may be paying for things you didn't know about.

How to Cancel Datadog

Log into app.datadoghq.com
Go to Organization Settings → Plan & Usage
Click "Downgrade to Free" or contact sales to cancel enterprise contracts
Remove the Datadog agent from all servers (it keeps sending data = charges)
Delete API keys to prevent accidental re-connection
For enterprise contracts: email your account manager — you may have a termination clause

⚠️ Critical: Remove agents FIRST. Even after "canceling," if the Datadog agent is still running on your servers, it continues sending metrics and you may be billed. Uninstall the agent from every host, container, and Lambda function.

Why Datadog Bills Spiral Out of Control

Per-host pricing + auto-scaling: Kubernetes spins up pods, each one is a "host" Datadog charges for
Custom metrics: $0.05/metric/month sounds cheap until you have 10,000 custom metrics ($500/month)
Log ingestion: $0.10/GB ingested. A chatty application logging 100GB/month = $10/month just for logs
Feature creep: Infrastructure + APM + Logs + Synthetics + RUM = 5 separate bills
No spending caps: Unlike AWS, Datadog has no built-in budget alerts (ironic for a monitoring tool)

Open-Source Alternatives (Save 80-90%)

Full Observability Stack

Grafana + Prometheus + Loki — The gold standard open-source stack. Metrics (Prometheus), logs (Loki), dashboards (Grafana). Self-hosted cost: just your server time
Grafana Cloud (free tier) — 10K metrics, 50GB logs, 50GB traces/month for free

APM / Tracing

Jaeger (free) — Distributed tracing, CNCF project, battle-tested at Uber
OpenTelemetry (free) — Vendor-neutral instrumentation, works with any backend
SigNoz (open-source) — Full Datadog alternative with a similar UI

Managed (Cheaper Than Datadog)

New Relic (free tier) — 100GB/month free ingestion, one full-access user
Betterstack ($29/mo) — Uptime monitoring + logs, clean UI
Axiom (free tier) — 500GB/month ingest, 30-day retention
Highlight.io (open-source) — Session replay + error tracking + logging

Migration Plan: Datadog → Grafana Stack

Week 1: Deploy Grafana + Prometheus. Install node_exporter on all hosts. Replicate your top 10 dashboards
Week 2: Add Loki for log aggregation. Migrate alert rules from Datadog monitors to Grafana alerting
Week 3: Add Jaeger or Tempo for distributed tracing. Instrument with OpenTelemetry
Week 4: Run both systems in parallel. Verify data matches. Remove Datadog agents

Bottom line: Datadog is a fantastic product with predatory pricing. The open-source Grafana/Prometheus/Loki stack does 90% of what Datadog does at 10% of the cost. For startups burning $10K+/month on monitoring, switching saves enough to hire another engineer. Just remember: remove the agents first.

Find More Subscriptions to Cancel →

⚡

Quick Cancel Guide: Datadog

Direct cancel link + step-by-step instructions

How to Cancel Datadog in 2026 (The $23/Host/Month Money Pit)