7 APM Tools Worth Evaluating in 2026

From Open Source to Enterprise – What Fits Your Stack

The APM landscape in 2026 spans a wider spectrum than ever – from fully open-source, self-hosted stacks to enterprise SaaS platforms with AI-driven root cause analysis. What’s changed most is the middle ground: OpenTelemetry has matured enough that teams can instrument once and choose (or change) their backend independently. That shift has made the APM decision less about which agent to install and more about which backend fits your operational model, budget, and data requirements.

For teams mid-migration to OpenTelemetry – or considering it – the choice of APM backend is more consequential than it used to be. Some vendors treat OTel as a first-class citizen; others accept OTel data but reclassify it in ways that increase cost. This guide evaluates seven APM platforms specifically through the lens of how well they fit modern, OTel-oriented infrastructure.

Cost estimates assume a mid-scale scenario: 30TB/month ingestion (~20TB logs, 7TB traces, 3TB metrics), 100 hosts, 20 full-platform users, 500,000 active metric series, and 30-day retention. Core observability only – no security, profiling, or synthetics. Estimates are directional, based on public rate cards as of early 2026.

Why Your APM’s OTel Strategy Matters

OpenTelemetry has become the de facto standard for instrumentation. The OTel Collector, SDKs, and semantic conventions are stable enough for production use. Most teams starting new instrumentation in 2026 are choosing OTel over proprietary agents. But not all APM backends treat OTel data equally:

• OTel-native platforms accept OTel data as the primary input format with no reclassification or surcharges. Your OTel metrics are just metrics.
• OTel-compatible platforms accept OTel data but may reclassify metrics as ‘custom metrics’ at premium rates, or support OTel traces but not OTel metrics natively. The distinction matters for cost and for how much of the OTel ecosystem you can actually use.
• Proprietary-first platforms have their own agents as the primary instrumentation path. OTel support exists but is secondary – fewer features, less documentation, sometimes higher cost.

For teams investing in OTel, the backend’s OTel stance directly affects both the migration path and the ongoing cost of running open-standard instrumentation.

1. CubeAPM

OTel stance: Native – built on OTel from day one

Overview

CubeAPM is a self-hosted, OTel-native observability platform covering APM, logs, infrastructure, Kubernetes, RUM, synthetic monitoring, Kafka monitoring, and error tracking. It was built on OpenTelemetry from day one – not retrofitted – so OTel data is the primary input format with no reclassification or premium pricing.

Recognized as a High Performer in G2’s Spring 2026 APM Grid Report. Used by brands like redBus, Delhivery ($3.5B valuation), Mamaearth ($1.2B valuation), Policybazaar, Practo, and others.

Core Strengths

• Consistently 70–75% lower cost than enterprise APM at scale
• Complete data ownership – no telemetry leaves your infrastructure
• OTel-native from day one – traces, metrics, and logs via OTel Collector
• No reclassification of OTel metrics as ‘custom metrics’ – all metrics priced equally
• Multi-agent compatible – accepts data from Prometheus, Datadog, New Relic, and Elastic agents alongside OTel. Teams can migrate incrementally without re-instrumenting everything at once.
• Data residency & Enterprise compliance: CubeAPM’s self-hosted architecture ensures full data residency control by design, while SOC 2 and ISO 27001 certification demonstrate adherence to industry-recognized security and governance standards.
• Unlimited retention, predictable pricing, no egress charges
• Engineering-level support that responds in minutes during incidents
• AI-based trace sampling included
• No proprietary agents to install, upgrade, or manage

Key Features

• Full-stack unified monitoring – APM, logs, infra, Kubernetes, Kafka, RUM, synthetics, error tracking
• Self-hosted/BYOC – data sovereignty by architecture
• AI-based trace sampling
• Unlimited retention, no egress charges
• Direct engineering support via shared channels

Pricing

$0.15/GB flat. No per-host, per-seat, or custom metrics fees.

At 30TB/month: ~$5,100/month all-in

Delhivery: 75% savings. Mamaearth: ~70% savings, migrated in under an hour. redBus: 4x faster dashboards, 50% faster MTTR.

Pros

• True OTel-native – no cost penalty for open standards instrumentation
• Lowest cost at scale in this list by a factor of 3-12x
• Incremental migration – run alongside existing agents without disruption
• Complete data ownership on your infrastructure

Cons

• Requires BYOC or on-prem deployment
• No autonomous anomaly detection
• SSO/RBAC less mature than enterprise incumbents

2. Datadog

OTel stance: Compatible, but with cost implications – OTel metrics often billed as custom metrics

Overview

Datadog is the market leader with 700+ integrations and a polished UI. It accepts OTel data, but the Datadog Agent remains the primary and best-supported instrumentation path. OTel metrics are frequently reclassified as custom metrics and billed at premium rates – a significant consideration for teams standardizing on OTel.

OTel Compatibility

• Accepts OTel traces, metrics, and logs via the OTel Collector or Datadog’s OTel-compatible endpoints
• OTel metrics are often classified as custom metrics – billed at up to $5 per 100 metrics/month beyond host allotment
• Datadog Agent remains the primary path for full feature access
• Some features (e.g., Network Performance Monitoring) require the Datadog Agent and don’t work with OTel-only instrumentation

Key Features

• Unified observability: metrics, logs, APM, RUM, synthetics, security
• 700+ integrations
• Watchdog AI for anomaly detection
• Service maps and deployment tracking

Pricing

Multi-dimensional: hosts + custom metrics + log ingestion + log indexing + APM spans + RUM. Custom metrics = 30-52% of bill at scale.

At 30TB/month: ~$30,000-$45,000+/month

Third-party calculators exist for modeling bills – use them before committing.

Pros

• Best integration ecosystem and UI
• Watchdog AI reduces alert noise
• Strong deployment tracking

Cons

• OTel metrics billed as custom metrics – penalizes open standards adoption (for teams where this is unacceptable, OTel-native platforms like CubeAPM or Grafana Cloud charge the same rate regardless of metric origin)
• Multi-dimensional billing compounds unpredictably
• No self-hosted option
• Full feature set requires Datadog Agent, not just OTel

3. Dynatrace

OTel stance: Compatible – OTel accepted alongside proprietary OneAgent

Overview

Dynatrace differentiates through Davis AI for causal root cause analysis. OTel support exists for traces, metrics, and logs, but OneAgent remains the primary instrumentation path and provides more features than OTel-only setups. Dynatrace Managed offers genuine self-hosted deployment.

OTel Compatibility

• Accepts OTel traces, metrics, and logs
• OneAgent is the primary instrumentation path – more features than OTel-only
• Smartscape topology mapping works best with OneAgent
• Teams can use OTel alongside OneAgent for specific services

Key Features

• Davis AI: causal root cause analysis
• Automatic discovery and Smartscape topology mapping
• Full-stack monitoring
• Dynatrace Managed: self-hosted deployment

Pricing

Consumption-based DPS, annual minimum (~$2,000/month). $0.08/hour per 8 GiB host, log ingest $0.20/GiB. 4 GiB minimum billing.

At 30TB/month: ~$20,000-$35,000+/month

Pros

• Best automated root cause analysis
• Genuine self-hosted option via Dynatrace Managed
• Automatic topology discovery

Cons

• Full value requires OneAgent – OTel-only setups miss some features
• Mandatory annual commitment
• 4 GiB minimum billing for small hosts
• Baselining period before Davis AI is fully effective

4. New Relic

OTel stance: Compatible – OTel is a supported instrumentation path with good documentation

Overview

New Relic has invested significantly in OTel support. NRDB accepts OTel data natively, and New Relic’s OTel documentation is among the best. No custom metrics penalty for OTel data. The trade-off is the dual pricing model (data + users) and SaaS-only deployment.

OTel Compatibility

• OTel traces, metrics, and logs accepted natively via OTLP endpoint
• No custom metrics penalty for OTel-originated metrics
• Good OTel documentation and examples
• New Relic agents still available but OTel is a well-supported alternative

Key Features

• NRDB: unified telemetry store
• NRQL: SQL-like querying
• Free tier: 100 GB/month + 1 user
• Distributed tracing, service maps

Pricing

$0.40/GB ingest + user fees ($49-$349/user/month). Data Plus at $0.60/GB for 90-day retention.

At 30TB/month: ~$20,000-$25,000+/month

Pros

• Good OTel support without custom metrics penalties
• NRQL is excellent for ad-hoc analysis
• Free tier removes evaluation friction

Cons

• Dual cost axes (data + users) both scale independently
• 8-day default retention
• No self-hosted option
• NRQL is proprietary – dashboards don’t port to other backends

5. Grafana Cloud (LGTM Stack)

OTel stance: Native – fully OTel-native with no proprietary agents

Overview

Grafana Cloud’s LGTM stack (Loki, Grafana, Tempo, Mimir) is one of the most OTel-aligned observability platforms available. No proprietary agents, no custom metrics penalty, and the open-source foundation means maximum portability. For teams that have committed to OTel as their instrumentation strategy, Grafana Cloud is a natural backend choice.

OTel Compatibility

• Fully OTel-native – designed around OTel data formats
• No custom metrics penalty – OTel metrics are just metrics
• No proprietary agents – OTel Collector is the instrumentation path
• Open-source components mean maximum portability if you switch backends

Key Features

• LGTM stack: Loki, Grafana, Tempo, Mimir
• Adaptive Metrics and Adaptive Logs for cost management
• Self-hosted (free) or managed cloud
• 13-month metrics, 30-day logs/traces on Pro

Pricing

$19/month base + usage. Logs ~$0.55/GB effective. Traces $0.50/GB. Metrics $8/1,000 series. Enterprise: $25K/year minimum.

At 30TB/month (managed): ~$15,000-$20,000+/month

Pros

• Fully OTel-native with no cost penalties
• Maximum portability – open-source components throughout
• Adaptive Metrics/Logs proactively reduce costs
• Self-hosted option at zero licensing cost

Cons

• Self-hosting at 30TB requires significant SRE investment
• Managed costs approach enterprise SaaS at high volumes
• No built-in AI/ML anomaly detection
• APM less mature than purpose-built tools

6. Elastic APM

OTel stance: Compatible – accepts OTel data via collector, but Elastic agents provide fuller integration

Overview

Elastic APM accepts OTel data via the OTel Collector writing to Elasticsearch. For teams migrating to OTel, Elastic can serve as a backend without requiring Elastic’s own agents – though the Elastic APM agents provide tighter integration with Elasticsearch features.

OTel Compatibility

• OTel Collector can write directly to Elasticsearch
• OTel traces and metrics are supported
• Elastic APM agents provide deeper integration than OTel-only
• No custom metrics penalty

Key Features

• Native Elasticsearch integration for log + trace correlation
• ML-based anomaly detection
• Self-hosted (free) or Elastic Cloud
• Service maps and distributed tracing

Pricing

Self-hosted free; Elastic Cloud deployment-based.

At 30TB/month (Elastic Cloud): ~$8,000-$15,000/month

Pros

• No custom metrics penalty for OTel data
• Zero incremental cost for existing ELK users
• Self-hosted keeps data on your infrastructure

Cons

• OTel-only setup misses some Elastic-agent-specific features
• Operational complexity at 30TB scale
• APM UX less polished than purpose-built tools like CubeAPM or Datadog
• SSPL licensing requires compliance review

7. Splunk Observability Cloud

OTel stance: Compatible – Splunk was an early OTel contributor, but the ecosystem is largely proprietary

Overview

Splunk was an early contributor to OpenTelemetry and accepts OTel data. However, the broader Splunk ecosystem – SIEM integration, query language, data format – is proprietary. OTel data flows in, but the value proposition is tied to the Splunk platform.

OTel Compatibility

• Accepts OTel traces and metrics
• Splunk was an early OTel project contributor
• Proprietary ecosystem around the OTel data once ingested
• Full-fidelity tracing works with OTel-instrumented services

Key Features

• Full-fidelity distributed tracing (no sampling)
• AI-based alerting
• Deep Splunk SIEM integration
• Real-time stream processing

Pricing

$15/host/month base. APM and logs via enterprise contract.

At 30TB/month: ~$35,000-$60,000+/month

Pros

• Full-fidelity traces with OTel instrumentation
• Early OTel contributor – good standards alignment
• Best security + observability integration

Cons

• Most expensive option at scale
• Proprietary ecosystem creates lock-in despite OTel ingest support
• Value primarily for existing Splunk organizations
• Significant deployment effort

Cost Comparison at 30TB/Month Ingestion

* OTel metrics in Datadog are often billed as custom metrics. All estimates use the reference scenario above. Vendor discounts can significantly reduce SaaS costs.

How to Choose

Choose CubeAPM if you want an OTel-native backend with no cost penalties, incremental migration from existing agents, and the lowest price at scale.

Choose Datadog if integration breadth matters more than OTel purity and your budget can absorb the custom metrics surcharge on OTel data.

Choose Dynatrace if automated root cause analysis is the priority and you’re comfortable running OneAgent alongside OTel instrumentation.

Choose New Relic if you want good OTel support with no custom metrics penalty and value NRQL for ad-hoc analysis.

Choose Grafana Cloud if maximum portability and zero proprietary lock-in are non-negotiable. The most OTel-aligned option with a managed service.

Choose Elastic APM if you already run ELK and want to route OTel data to Elasticsearch without adding a new backend.

Choose Splunk if your organization has existing Splunk infrastructure and values the security integration, with OTel as the instrumentation layer.

Final Thoughts

OpenTelemetry has fundamentally changed the APM decision. When your instrumentation is vendor-neutral, the backend becomes a replaceable component – which means the evaluation criteria shift from ‘which agent do we install’ to ‘which backend gives us the best cost, operational model, and data control for our OTel data.’

The platforms that treat OTel as a first-class citizen – no reclassification, no surcharges, no feature gaps – give teams the most flexibility going forward. The platforms that accept OTel but price or feature-gate it differently create a hidden cost of adopting open standards. Both can be the right choice depending on what else the platform offers, but the distinction is worth understanding before committing.

For teams early in their OTel migration, the practical advice is: instrument with OTel, run a proof-of-concept with two or three backends, and compare the actual bill at realistic data volumes. The sticker price tells one story; the invoice after a month of production data tells another.

Keywords: OpenTelemetry APM tools 2026, OTel-native observability, APM for OTel migration, open source APM, Datadog OTel custom metrics, self-hosted APM, observability platform comparison, CubeAPM