Pricing Overview
Vector is a fully open-source observability data pipeline released under the Mozilla Public License v2.0 and maintained by Datadog. The software itself is completely free -- there are no license fees, no per-node charges, and no usage caps baked into the tool. Vector ships as a single Rust binary with zero runtime dependencies, which means the real cost of running it comes down to the compute, memory, and storage your infrastructure requires to handle your observability data volume.
We find this model compelling for teams that want full control over their log and metric routing without committing to a vendor's pricing structure. Vector can run as a daemon on individual hosts, as a sidecar alongside application containers, or as a centralized aggregator -- each deployment pattern carries different operational cost profiles.
Plan Comparison
Because Vector is open-source infrastructure software, it does not have traditional pricing tiers. Instead, costs depend on how you deploy it and what destinations you route data to. We break the options into three practical deployment paths.
| Deployment Path | Software Cost | Infrastructure Requirements | Operational Overhead | Best For |
|---|---|---|---|---|
| Daemon (per-host agent) | $0 | Low -- ~50-100 MB RAM per host | Minimal -- runs alongside existing workloads | Teams collecting logs/metrics from individual servers |
| Sidecar (per-container) | $0 | Very low -- ~20-50 MB RAM per pod | Low -- managed via orchestration (Kubernetes) | Containerized microservice architectures |
| Aggregator (centralized) | $0 | Moderate -- dedicated instances, 2-8 GB RAM depending on throughput | Higher -- requires capacity planning and monitoring | High-volume environments routing to multiple sinks |
The daemon deployment is the simplest path. You install Vector on each host and configure it to collect and forward logs and metrics. The sidecar model works well in Kubernetes, where Vector runs as a DaemonSet or sidecar container. The aggregator model is the most powerful -- it receives data from multiple sources, applies transforms using Vector Remap Language (VRL), and routes to any combination of Vector's 61 supported sinks.
All three paths use the same binary and the same configuration format (YAML, TOML, or JSON). The only cost difference is how much compute you allocate.
Hidden Costs and Considerations
While Vector itself is free, several cost factors deserve careful planning. Destination costs are the biggest variable -- if you route logs to Datadog, Elasticsearch, or AWS S3, those services charge based on ingestion volume, storage, and query patterns. Vector's transform capabilities (filtering, sampling, aggregation) can significantly reduce downstream costs by dropping unnecessary data before it reaches expensive sinks.
Compute costs scale with throughput. Vector handles roughly 10 TB/day per vCPU for simple routing, but complex VRL transforms reduce that throughput. Teams running heavy transformations on high-volume pipelines need to budget for additional CPU capacity. Memory usage increases with buffer sizes -- if you configure disk-based buffers for durability, add SSD storage costs to the equation.
Operational expertise is the hidden cost that most teams underestimate. You need engineers who can write VRL transforms, design pipeline topologies, monitor Vector's own health metrics, and handle capacity planning. There is no vendor support hotline unless you use Vector through Datadog's commercial offering.
Cost Estimates
We can estimate realistic operational costs based on common deployment scenarios. These figures assume AWS infrastructure pricing.
| Scenario | Daily Volume | Infrastructure | Estimated Monthly Cost |
|---|---|---|---|
| Small team (daemon mode) | 10 GB/day logs | 5 t3.small instances (shared workload) | $45-75 (Vector's share of compute) |
| Mid-size (aggregator + daemons) | 500 GB/day logs + metrics | 2 c5.xlarge aggregators + 50 daemon hosts | $350-600 (aggregator compute + network) |
| Large-scale (multi-tier) | 5 TB/day logs + metrics | 8 c5.2xlarge aggregators + 200 daemon hosts | $2,000-3,500 (compute + disk buffers) |
These estimates cover only Vector's infrastructure footprint. The destination services (Datadog, Elasticsearch, S3, etc.) add their own charges, which typically dwarf the pipeline cost itself. Vector's filtering and sampling transforms are your primary lever for controlling those downstream bills.
How Vector Pricing Compares
Vector competes in the observability pipeline space against both open-source alternatives and commercial platforms. The comparison below puts Vector's cost model in context.
| Tool | Pricing Model | Starting Price | Free Tier | Open Source |
|---|---|---|---|---|
| Vector | Open Source | $0 | Full project is free | Yes (MPL 2.0) |
| New Relic | Usage-Based | $0 (free tier) / $19/mo per host | Yes (100 GB/month ingest) | No |
| Observe | Usage-Based | $0.49/GB (logs) | Limited free tier | No |
| Elastic Observability | Paid | $95/mo (Standard) | No (self-managed Elasticsearch is open-source) | Partial |
Vector and the commercial platforms serve different roles. Vector is a data pipeline -- it collects, transforms, and routes observability data but does not store or query it. New Relic, Observe, and Elastic Observability are full observability platforms with storage, querying, dashboards, and alerting included.
The cost advantage of Vector becomes clear when paired with cost-effective storage backends. Routing logs through Vector to S3 at $0.023/GB/month is dramatically cheaper than ingesting directly into New Relic at $0.30/GB or Observe at $0.49/GB. For a team processing 1 TB/day, the difference between routing to S3 versus a commercial observability platform can exceed $8,000/month.
New Relic offers the most generous free tier at 100 GB/month of ingest, making it attractive for small teams. Elastic Observability starts at $95/month for managed deployments but can be self-hosted using the open-source Elasticsearch stack. Observe targets teams that want a managed observability platform with usage-based pricing but no free self-hosted option.
Vector's strongest value proposition is vendor neutrality. With 47 sources and 61 sinks, you can route data to any combination of destinations and switch backends without modifying your collection infrastructure. That flexibility translates directly into cost optimization opportunities that locked-in platforms cannot match.