Five myths about Sentry’s 2026 error-tracking pricing debunked

The volatility of usage-based billing

Datadog users often report billing surprises that exceed their budgets by 3x to 10x. Sentry uses an event-based model where costs scale with errors, transactions, replays, logs, and attachments. Datadog uses a high-water mark system. If a team scales from 50 to 200 hosts during a five-day product launch, Datadog bills that entire month at the 200-host rate. This creates a massive discrepancy between average usage and the final invoice.

Datadog’s custom metric charges add unexpected weight.

Datadog charges $0.10 per GB to collect logs and then $1.70 per million events to index them. Custom metrics also incur surcharges, and metrics sent via OpenTelemetry fall into this expensive bucket. If you add a tag with many unique values, cardinality can explode your bill. A single metric with 1,000 unique tag values can generate $4,500 in monthly overages. Datadog’s per-host APM fees of $31 to $40 per month land on top of mandatory infrastructure fees, meaning every new service deployment increases the bill in a way that scales with your footprint rather than your actual error volume. Datadog also charges for LLM Observability, where token ingestion costs approximately $0.10 per 1,000 tokens.

Feature depth and user costs

Sentry’s debugging power justifies the event volume. The Seer Agent, available in 2026, lets users describe a symptom in plain language to find the upstream cause. Datadog error tracking requires APM or RUM spend to exist. Sentry includes session replay and profiling within its existing debugging loop. Sentry’s logs arrived in 2025.

Sentry handles users differently.

The Sentry Team plan includes 20 users and costs $26 per month for annual billing. New Relic and Datadog often charge based on user seats or host counts. New Relic’s user-based pricing can become a significant portion of the monthly bill. New Relic’s free forever plan includes 100GB of data ingestion per month. Additional New Relic data ingestion costs $0.3 per GB or $0.5 per GB. Sentry’s Team plan includes 50,000 errors and 100,000 transactions. Additional errors on the Business tier cost between $0.0011125 and $0.0003000 per error. Sentry also allows users to add $40 per month per active code contributor for Seer. Sentry’s developer tier provides 5,000 errors and 10,000 performance units per month for free.

Scale and functional capacity

Sentry covers the full application-layer debugging loop. It provides stack traces, breadcrumbs, session replay, and continuous profiling. Many teams think Sentry lacks the depth of a full observability platform. Sentry’s features link errors directly to the commit that caused them. You should check your own usage before assuming Sentry lacks capability.

Sentry handles scale via events.

Sentry’s session replays cost roughly $0.006 per replay on the Team plan. Log storage beyond the included 5GB costs $0.50 per GB. Attachment storage for debug files and source maps costs $0.15 per GB per month. If you need to manage high volume, does Sentry’s per-event model create more stress than a flat-rate alternative? Sentry’s error grouping uses a machine learning model to prevent 20% more duplicate issues. A single noisy deploy can spike Sentry costs, as one team reported a 3.2x bill spike after hitting 1.4 million events in a single day. Sentry’s Team plan includes 5 million spans. Sentry’s Business tier offers custom data retention from 90 days to over a year. Sentry’s Team plan provides 90-day event retention.