Datadog Alternative

Gigapipe is the open source, OpenTelemetry-native alternative to Datadog. It is priced by infrastructure instead of ingestion, takes the agents you already run, and keeps your data in storage you own.
Why teams leave Datadog
Three things that are true of every consumption-priced, closed-format observability vendor.

Your data, their format

Telemetry lands in a format only Datadog can read, and every dashboard, monitor and saved query is written against a syntax that exists in one product. The longer you stay, the more leaving costs.

A bill you cannot predict

Each signal is priced separately and measured by an accounting only the vendor can see. Send more logs this month and less next month, and the invoice moves under you.

Priced to make you look away

When observability bills by consumption, the rational move mid-incident is to sample traces and drop logs. The data gets cut exactly when it is worth the most.

Run both. Decide afterwards.
Gigapipe accepts telemetry sent from the Datadog Agent natively. Send us your data, compare it against Datadog, with no commitment and no extra cost.

Nothing to rip out

Use the same Datadog Agent to send telemetry to both Datadog and Gigapipe. No architecture changes required.

No commitment

Run Gigapipe OSS on your own hardware for as long as you need. No trial window, no sales call, nothing to sign.

Your data. Your rules.

Sending the same telemetry twice costs you nothing extra. Full stack observability, without your data being held hostage.

Gigapipe compared with Datadog
Data format
Open schema in storage you own, readable with plain SQL.
Proprietary, reachable only through Datadog.
What you pay for
OSS is free, always. Cloud is the machine in your contract, with ingestion and queries unlimited within it.
Per GB, per host, per seat, per custom metric. Indexed spans extra.
Changing your signal mix
Free. Mostly logs today, mostly metrics tomorrow, same price.
Each signal is its own line item, so the invoice moves.
Storage accounting
Known capacity, and you use all of it. Nothing to estimate.
Priced per signal on the vendor's own measurement.
Licence
OSS is AGPLv3. Cloud is that engine plus hosting, support and AI features.
Proprietary and closed source.
Where data lives
Your own database and object storage.
Datadog-operated cloud.
Query languages
LogQL, PromQL, TraceQL and SQL.
Proprietary syntax, one UI.
Visualisation
Grafana or Perses, built-in datasources, no plugin.
Datadog UI only.
OpenTelemetry
Native OTLP over gRPC and HTTP, for all four signals.
Accepted, then remapped into their schema.
Exit cost
A data copy.
A negotiation. Historical data stays behind.
See the current plans
Keep the Datadog Agent
Gigapipe implements the Datadog API natively, so you can ingest telemetry with the Datadog Agent you already run. You change the exporting address, nothing else. Nothing to re-instrument, nothing to re-architect.
datadog.yaml
# Verified against gcr.io/datadoghq/agent:7 and Gigapipe on :3100.
dd_url: http://<your-gigapipe-host>:3100
skip_ssl_validation: true

# Agent 7 compresses with zstd by default, which Gigapipe does not
# decode yet. Pin the compressors until it does.
serializer_compressor_kind: zlib

logs_enabled: true
logs_config:
  logs_dd_url: <your-gigapipe-host>:3100
  logs_no_ssl: true
  use_http: true
  compression_kind: gzip

# Traces leave over OTLP.
otlp_config:
  receiver:
    protocols:
      grpc:
        endpoint: 0.0.0.0:4317
LogsPOST /api/v2/logs

Datadog log intake, accepted unchanged.

MetricsPOST /api/v2/series

Datadog metrics intake, accepted unchanged.

TracesPOST /v1/traces

OTLP over HTTP, or gRPC on port 4317.

ProfilesPOST /v1development/profiles

OTLP profiles, the fourth signal.

The simplest migration
1

Run Gigapipe

Gigapipe runs in whatever topology you already have: VMs, bare metal, Kubernetes, containers. If it can run a Go binary, it can run Gigapipe. We match your architecture instead of imposing one.

2

Send to both

Keep your Datadog Agent, repoint it at Gigapipe, and compare the two side by side on live traffic.

3

Rebuild the dashboards

Rebuild them once in Grafana or Perses. Hard work, done a single time, and then they are yours to keep. Gigapipe speaks the industry standards.

4

Cut over when you are ready

Once the data lines up and you are comfortable, swap to the industry standard tool. No more per-event pricing you cannot estimate. Just a fixed price based on machine specs.

Datadog alternative FAQ
Can I keep using the Datadog Agent?
Can I run Gigapipe alongside Datadog before switching?
How does Gigapipe pricing differ from Datadog?
Am I locked into Gigapipe instead?
Can I self-host it?
Do I have to give up my dashboards?
Is it OpenTelemetry-native?
Ready for a bill you can predict?