Why is Datadog so expensive?

Datadog charges separately for ingestion ($0.10/GB), log indexing ($1.06–2.50 per million events by retention window — ingest-only logs are not searchable), and infrastructure hosts ($15+/host/month). At 10 GB/day with 30-day retention and 10 hosts, list pricing adds up to roughly $690/month before any other product.

What is the cheapest Datadog alternative for logs?

At typical SMB volumes, object-storage-backed platforms like Axiom ($25/mo including 1 TB ingest) and Better Stack ($25/mo bundles with 30-day retention) price an order of magnitude below Datadog for equivalent searchable-log workloads.

OBSERVABILITY · PRICING COMPARE

Observability pricing compare prices verified 2026-08-24

Capture-dated list prices, ranked on your actual volume. Quote-driven platforms are shown but never guessed onto the chart.

Logs ingested (GB/day) Retention window (days) Hosts monitored

Bill-shock check: at 10 GB/day × 30d × 10 hosts, Datadog list pricing runs 27.6× the cheapest ranked alternative ($690.40 vs $25.00/mo).

The indexed-events line is the shock: ingest-only logs are not searchable, so querying anything adds $1.06–2.50 per million events per month on top of ingestion.

#	Platform	Ingest /mo	Retention /mo	Hosts /mo	Total /mo	× cheapest
6	Datadog	$30.40	$510.00	$150.00	$690.40	27.6×
3	New Relic	$81.60	incl.	—	$81.60	3.3×
4	Grafana Cloud	$114.30	incl.	$72.00	$205.30	8.2×
2	Better Stack	$26.40	incl.	—	$51.40	2.1×
1	Axiom	$0.00	incl.	—	$25.00	cheapest
5	Dynatrace	$60.80	$4.00	$290.00	$354.80	14.2×

Not ranked — and why that's honest

Sumo Logic — NOT RANKED: no public flat $/GB — official estimator implies an effective $2.05–3.14/GB depending on scan profile (US region; +10–20% other regions), i.e. quote-driven pricing that cannot be honestly modeled without your scan profile. Retention up to 365 days customer-defined. No perpetual free tier (30-day trial).
Honeycomb — NOT RANKED: priced per EVENT (Pro from $150/mo for 50M events ≈ $3/M events, 60-day retention, unlimited seats) — a $/GB normalization requires assuming an average event size we refuse to invent. Generous free tier: 20M events + 100M datapoints/mo. Best suited to high-cardinality debugging, not bulk log archival.
Sentry — NOT RANKED: error/crash tracking first (Team $26/mo base, quota-based errors/spans/logs; logs add-on $0.50/GB) — its 5 GB/mo included log volume shows the product is not aiming to be your log pipeline. Included for completeness of the observability-bill conversation, not compared on $/GB.

data.csv data.json

Methodology

All rates are published list prices fetched from official pricing pages on 2026-08-24; every row links its source.
Model: ingest = GB/day × 30.4 × rate (free allowances deducted); retention bills only days beyond each platform's included window; hosts and platform fees add on top.
Datadog's indexing is officially priced per million events; we use the standard ≈1M events ≈ 1 GB approximation and label it. We do not apply the same trick in reverse: event-priced platforms (Honeycomb, Sentry) and quote-based platforms (Sumo Logic) stay unranked.
Excluded line items are stated per vendor — e.g. New Relic full-platform user seats ($99–349/user/mo), which can exceed the usage bill.