Top 10 Datadog Alternatives You Should Know About in 2026

If you’ve watched a Datadog renewal quote climb three years running, 2026 is the first year in a while where the shortlist is worth real attention. A wave of ingestion-based and customer-owned-storage products has matured into production tooling, and OpenTelemetry adoption lets you fan out telemetry to two backends in parallel without rewriting instrumentation.

This guide covers why teams are leaving Datadog in 2026, the five filters that separate a real upgrade from a lateral move, and 10 alternatives worth a proof of concept.

Why Teams Are Switching From Datadog in 2026

Datadog renewals break down on billing mechanics, not product capability. Infrastructure, application performance monitoring (APM), logs, and custom metrics meter in ways a year-one quote rarely shows. Four pain categories surface together:

• Per-host billing compounds in autoscaling Kubernetes: Infrastructure monitoring starts at $15 per host per month on the Pro tier (annual), so a Horizontal Pod Autoscaler spinning up 200 ephemeral nodes adds line items teams report they didn’t model. Coralogix’s per-gigabyte meter takes host count out of the equation.
• Log ingest and indexing meter separately: Logs run $0.10 per gigabyte for ingest plus $2.50 to $3.75 per million events for 30-day indexing (annual to on-demand, verified June 2026), so the bill grows at rest. The Coralogix TCO Optimizer routes logs into Frequent Search, Monitoring, Compliance, and Blocked pipelines based on policies you define for each data stream.
• Modular SKU stacking: APM, real user monitoring (RUM), serverless, and custom metrics each bill independently, with Fargate APM at $2.60 to $3.70 per task per month (annual to on-demand, verified June 2026). Coralogix bundles logs, metrics, traces, RUM, and security on one meter.
• Proprietary agent and query layer slow migrations: The Datadog Agent and Datadog Query Language create a switching cost an OpenTelemetry (OTel) collector alone can’t dissolve. DataPrime and OpenTelemetry Protocol (OTLP) ingestion cut over without rewriting the collection layer.

Teams running these evaluations want a billing model that doesn’t penalize the architecture they already run.

What to Look for in a Datadog Alternative

The strongest evaluations test a candidate against your real telemetry, not a vendor matrix. Project spend at two, five, and 10 times current volume and weigh each filter against the billing dimension that hurts most today. Five filters separate an upgrade from a lateral move:

• Billing dimension and total cost of ownership: Pricing should scale with data volume, not with hosts, containers, or users.
• OpenTelemetry-native ingestion: OTLP should land as a first-class path on the Coralogix OpenTelemetry docs and equivalent vendor pages, with semantic conventions preserved.
• Cross-signal coverage in one product: Logs, metrics, traces, RUM, and security information and event management (SIEM) signals should investigate together so a metric anomaly walks straight to the correlated traces and logs.
• Customer-owned storage: Object-storage rates with no rehydration fees separate retention cost from query cost.
• Deployment flexibility: Self-hosted, bring-your-own-cloud (BYOC), and air-gapped paths cover workloads General Data Protection Regulation (GDPR) rules out for SaaS-only tools.

These filters narrow the field faster than any feature checklist.

The 10 Datadog Alternatives Worth Evaluating in 2026

Most teams leaving Datadog shortlist from three buckets: cross-stack observability on one ingestion meter, cloud-native SIEM bundled with observability, and open-source stacks bought as a managed service or self-hosted. The table maps each tool to starting price, deployment, signal coverage, and OTel support. Every price below links to the vendor’s current pricing page and was verified in June 2026; vendor pricing changes, so check the linked page before a renewal decision.

Platform	Starting Price	Deployment	Signal Coverage	OTel Support
Coralogix	$0.42/GB logs, $0.05/GB metrics, $0.16/GB traces	SaaS with customer-owned S3	Logs, metrics, traces, SIEM, RUM, AI	OTel-native, OpAMP
New Relic	Free 100 GB/month, then $0.40/GB	SaaS only	APM, logs, metrics, traces, RUM	OTel first-class
Dynatrace	$0.08/hour per 8-GiB host (Full Stack DPS)	SaaS, Dynatrace Managed	APM, logs, metrics, traces, RUM, security	OTel accepted, OneAgent default
Grafana Cloud	Free tier; Pro from $19/month + usage	SaaS or self-host the LGTM stack	Logs, metrics, traces, profiles, RUM	Native via Alloy
Elastic Observability	$0.07/GB Serverless Logs Essentials	SaaS, Serverless, self-managed	Logs, metrics, traces, RUM, security	Native via EDOT
Splunk Observability	From $15/host/month (annual)	SaaS only	Metrics, traces, RUM, synthetics; logs via Splunk Cloud	OTel-native (default ingestion)
Honeycomb	From $130/month for 100M events	SaaS, Private Cloud (Enterprise)	Traces, events, metrics, frontend (Enterprise)	OTel-native
SigNoz	$49/month Cloud Teams; free self-hosted (MIT)	SaaS, BYOC, self-hosted	Logs, metrics, traces, infra	OTel-only ingestion
Sumo Logic	Flex credit packs (quote-based); free 500 MB/day tier	SaaS only	Logs, metrics, traces, RUM, SIEM, SOAR	OTel Collector distribution
Logz.io	Per-GB consumption (quote-based); 14-day trial only	SaaS only	Logs, metrics, traces, SIEM	OTel-native (OTLP receivers)

1. Coralogix: In-Stream Observability with Customer-Owned Storage

Coralogix runs logs, metrics, traces, and security data through one OTel-native pipeline, where Datadog meters each product separately. Streama analyzes, anomaly-baselines, and alerts on data in flight, then writes it to your own Amazon Simple Storage Service (S3) bucket in open Parquet. Olly, Coralogix’s autonomous observability agent, cross-references telemetry with a connected GitHub repo to surface root cause and point you to the code path behind an incident.

Pros:

• Ingestion-based pricing at $0.42/GB logs, $0.05/GB metrics, and $0.16/GB traces, with every feature on every plan, no per-host or per-user fees, and 24/7 support at a 17-second median response time.
• Customer-owned indexless storage in open Parquet, so historical queries skip reindexing and rehydration fees.
• Named a Visionary in the 2025 Gartner Magic Quadrant for Observability Platforms, its first year in the report.

Cons:

• Shorter track record than the longest-tenured Datadog incumbents in some enterprise procurement cycles.
• Index-first teams need a brief ramp on in-stream concepts and DataPrime pipe syntax, though Olly’s natural language path closes the gap.

Best for: Teams leaving Datadog over per-host billing or agent lock-in wanting one meter across observability, security, and AI with data in their own bucket.

2. New Relic

New Relic prices SaaS observability across per-gigabyte ingest and per-user seats, with a perpetual free tier covering 100 GB monthly and one Full Platform user.

Pros:

• The free tier covers a real evaluation without a contract or credit card.
• No per-host charges keeps autoscaling Kubernetes footprints predictable.

Cons:

• Per-user pricing compounds past 20 engineers, where Coralogix has no per-user fees.
• Data Plus doubles the per-gigabyte rate for HIPAA and FedRAMP features.

Best for: APM-led teams under 20 engineers that want a transparent free tier.

3. Dynatrace

Dynatrace runs on OneAgent for auto-instrumentation and Davis causal AI for root cause, with telemetry flowing into the Grail lakehouse. The Dynatrace Platform Subscription (DPS) bills by capacity-hour against an 8-GiB host at roughly $0.08 per hour.

Pros:

• Root cause findings tie to the live service graph, shortening investigation for supported runtimes.
• DPS folds OTel metrics and traces into the cross-stack subscription at no extra meter.

Cons:

• OneAgent bytecode injection creates soft lock-in even when OTel runs alongside it, where Coralogix stays OTel-native end to end.
• Kubernetes Platform Monitoring bills per pod-hour, compounding in dense microservice environments.

Best for: Enterprise ops teams where causal root cause on the live service graph is the deciding criterion.

4. Grafana Cloud

Grafana Cloud runs the open-source Loki, Grafana, Tempo, Mimir (LGTM) stack plus Pyroscope and Alloy for OTLP-compatible collection. Paid usage starts at $19 per month on Pro plus per-signal rates, with a free tier for evaluation.

Pros:

• Loki, Mimir, Tempo, and Pyroscope ship under Apache 2.0, so the same backends run free in self-host.
• Native OTel ingestion through Alloy fans out to other backends during overlap, making a Datadog-to-Grafana migration reversible.

Cons:

• Three-component pricing across Process, Write, and Retain is hard to estimate before usage data arrives, where Coralogix bills one rate per signal.
• Active-series counts and cardinality drift swing the bill month over month, where the Coralogix TCO Optimizer holds cost predictable on policies you define across Frequent Search, Monitoring, Compliance, and Blocked pipelines.

Best for: Site reliability engineering (SRE) teams invested in Grafana dashboards who want managed convenience without giving up open-source portability.

5. Elastic Observability

Elastic Observability runs on Elasticsearch across self-managed, Cloud Hosted, and Serverless paths. Elastic Distributions of OpenTelemetry (EDOT) hit general availability in 2025, and Serverless pricing starts at $0.07 per gigabyte on Logs Essentials.

Pros:

• Deployment range spans SaaS Serverless to self-managed clusters, fitting regulated estates SaaS-only products rule out.
• EDOT gives you a real OTel-native path, with Open Agent Management Protocol (OpAMP) handling fleet config.

Cons:

• Data is indexed before query, compounding retention cost at high volume, where Coralogix queries open Parquet archives directly with no rehydration fee.
• Self-managed Elasticsearch needs platform engineers for shard management and high availability.

Best for: Engineering teams with Elasticsearch operators on staff needing the broadest deployment range.

6. Splunk Observability: OTel-Native Tracing Under Cisco

Splunk Observability Cloud sits inside the Cisco estate after the 2024 acquisition. Its architectural call is treating OpenTelemetry as the default ingestion path, and NoSample tracing retains 100 percent of spans without head- or tail-sampling.

Pros:

• OTel as the default collector path keeps instrumentation portable if you fan out to a second backend.
• NoSample full-fidelity tracing captures every span, keeping tail-latency outliers visible.

Cons:

• Log ingestion routes through Splunk Cloud or Splunk Enterprise, dragging a second SKU into the bill.
• Per-host bundles for infrastructure, APM, and synthetics meter separately, where Coralogix bills per gigabyte on one meter.

Best for: Security-aligned enterprises running Splunk Enterprise Security who want OTel-native tracing under one Cisco contract.

7. Honeycomb: Event-Shaped Debugging on a Columnar Backend

Honeycomb stores every signal as a wide structured event in its Retriever columnar store, which lets you query high-cardinality fields like user_id or tenant_id without pre-defined indexes. Pro pricing starts at $130 per month for 100 million events.

Pros:

• Retriever’s columnar layout handles high-cardinality querying without the pre-aggregation tax metrics-first backends charge.
• BubbleUp surfaces the dimensions where anomalous events cluster, shortening “what changed?” investigations.

Cons:

• Infrastructure and SIEM aren’t in the product, so cross-stack consolidation needs a second vendor, where Coralogix bundles them on one meter.
• Service Map, Frontend Observability, and Private Cloud sit behind the Enterprise tier.

Best for: Microservices teams whose debugging hinges on request-level context.

8. SigNoz: OTel-Native Open Source on ClickHouse

SigNoz lands logs, metrics, and traces in one ClickHouse columnar backend with OpenTelemetry as the only instrumentation path, so one query model covers all three signals. The self-hosted core is MIT-licensed, and the Cloud Teams plan starts at $49 per month.

Pros:

• MIT licensing lowers lock-in risk for teams wanting a self-hosted path with no per-seat or per-host fees.
• One ClickHouse backend covers all three signals under one query model, with OTel-only ingestion.

Cons:

• ClickHouse operations fall on the team in self-hosted deployments, where Coralogix is fully managed.
• Default cloud retention sits at 15 days, and finer governance is still maturing.

Best for: OTel-native teams wanting a permissive license and one backend across three signals.

9. Sumo Logic: Converged Observability and Security on Flex Credits

Sumo Logic ships log analytics, APM, Cloud SIEM, and security orchestration, automation, and response (SOAR) under one contract. The Flex Licensing model shifts cost from raw ingestion to credits covering analytics, retention, and security together.

Pros:

• Cloud SIEM, user and entity behavior analytics (UEBA), and SOAR share one pipeline with logs and traces.
• Federal Risk and Authorization Management Program (FedRAMP) Moderate authorization keeps federal workloads under one contract.

Cons:

• Per-credit dollar rates aren’t published, so cost modeling needs a sales conversation.
• Credit math rewards steady workloads and taxes bursty query volume, where Coralogix prices on ingestion only.

Best for: Mid-market and federal teams wanting observability, SIEM, and SOAR under one contract.

10. Logz.io: Managed Open Source Through Open 360

Logz.io wraps OpenSearch, Prometheus-compatible storage, and Jaeger into a managed service positioned as an upgrade off self-hosted ELK or Prometheus. The Open 360 platform ties the three backends together with monthly capacity reallocation across signal types.

Pros:

• OpenSearch, Prometheus, and Jaeger keep query languages portable across vendors.
• Monthly capacity reallocation means a logs-heavy quarter doesn’t strand paid-for metrics budget.

Cons:

• No permanent free tier past the 14-day trial, with per-gigabyte consumption scaling hard on log volume.
• OpenSearch indexes before query, so high-retention workloads pay an indexing cost per gigabyte, where Coralogix writes open Parquet to your own S3 bucket.

Best for: Teams migrating off self-hosted ELK or Prometheus wanting managed operations on familiar backends.

How to Choose the Right Datadog Alternative for Your Team

Narrow the field by ranking the Datadog billing dimension that hits hardest, then match it to the architecture that neutralizes it. Four checks stress-test that match against your own data:

• Audit your highest-cost Datadog SKUs first: Pull the last three invoices and rank Infrastructure, APM, Logs, and Custom Metrics by spend. The top SKU dictates the trade-off you can’t afford to repeat.
• Project costs at two, five, and 10 times current volume: Model retention, user seats, and alert quotas separately from the headline ingest rate.
• Run an OpenTelemetry Collector in fan-out mode: Send the same telemetry to Datadog and the candidate during overlap, comparing dashboards, alert latency, and cost behavior on real traffic.
• Plan the dashboard and alert migration as a real line item: Naming changes often require updating queries, dashboards, and alerting rules.

Those four checks usually leave two candidates and a defensible answer for the renewal call.

Test a Datadog Alternative on Your Own Production Telemetry

You only see how a tool handles your stack once peak ingest, retention windows, and a live on-call rotation run through it. The gap between a controlled demo and the third invoice is where most migrations come undone.

If per-host billing or log indexing costs are pushing you off Datadog, try a free Coralogix trial and route real production data through ingestion-based pricing for 14 days. You’ll see what your Datadog ingest costs on one meter with no host or query axis on top.

Frequently Asked Questions About Datadog Alternatives

How do you choose the right Datadog alternative?

It depends on which Datadog pain category drove the evaluation. Teams leaving over per-host scaling and log indexing costs tend to shortlist Coralogix for ingestion-based pricing with customer-owned S3 storage. Grafana Cloud fits teams invested in the LGTM stack, and Honeycomb suits event-level debugging for high-cardinality microservices.

Why is Datadog so expensive?

Datadog bills each capability as a separate SKU, so Infrastructure, APM, Logs, RUM, and Custom Metrics each carry their own meter. Per-host billing scales with peak node count, log indexing runs $2.50 to $3.75 per million events, and custom metrics cardinality multiplies the bill once high-cardinality tags enter the data model. Coralogix runs every signal on one ingestion-based meter with no per-host or per-query fees.

Is there a free or open-source alternative to Datadog?

SigNoz ships an MIT-licensed self-hosted core covering logs, metrics, and traces on ClickHouse, with enterprise features under a separate license. Grafana Cloud’s free tier covers 50 GB of logs, 10,000 active series, and 50 GB of traces (verified June 2026), and Honeycomb runs 20 million events per month free. Coralogix offers a 14-day free trial for teams that want to test ingestion-based pricing on real production telemetry.

Can you self-host Datadog?

As of June 2026, Datadog CloudPrem covers bring-your-own-cloud log management in your own infrastructure, and Datadog has not announced full self-hosted observability across traces and metrics. SigNoz, Elastic self-managed, and the Grafana LGTM stack all offer fully self-hosted paths. Coralogix processes telemetry as SaaS while writing the data to your own S3 bucket in open Parquet.