Pydantic Logfire Enterprise Plans
Pydantic Logfire has two enterprise options, depending on how much control you want:
- Enterprise Cloud – Fully managed. SLA-backed. No infrastructure work.
- Enterprise Self-Hosted – Run it yourself, in your own Kubernetes cluster. Same product, but with full data control.
Enterprise Cloud
For teams who want managed infra, advanced compliance, guaranteed uptime, and top-tier support
Good for:
- Longer retention than the standard 30 days
- SLA requirements
- Custom DPAs, HIPAA BAAs
- Invoice billing
- Custom SSO (Okta, Azure Entra ID, etc.)
Features
| Feature | Description |
|---|---|
| Fully Hosted | We run everything in our cloud. Upgrading from SaaS Pro takes 5 mins, no migration required. |
| 24/7 Support | You get priority access and support |
| Compliance & Billing | DPAs, BAAs, invoices, whatever your legal team needs. |
| Custom Retention | Store data for longer. You choose how long. |
| Custom SSO | Plug in your identity provider, Dex makes it work with almost anything. |
We use Dex, which works well with Okta, Azure Entra ID, Google Workspace, Auth0, LDAP, and anything else OIDC/SAML.
Enterprise Self-Hosted
For teams who want full control over their data.
Good for:
- Adhering to strict compliance or data residency requirements
- Teams with Kubernetes know-how
Features
| Feature | Description |
|---|---|
| All Enterprise Features | Custom data retention & SSO are configurable, 24/7 support. |
| Open-Source Helm Chart | You are free to explore the Helm chart code. |
| Uncomplicated Infrastructure | The only stateful services are Postgres and Object Storage. Use any S3-compatible backend. Managed Postgres (e.g. RDS) is fine. |
| Scalability | The system is designed to autoscale. |
Support Includes
- Help getting set up
- Advice for production best practices
- Real engineers for troubleshooting
- Optional health checks and performance tuning
A Look Under the Hood
In addition to the open-source Helm chart, our self-hosted documentation contains recommended configurations, how-tos for running locally with Tilt and troubleshooting guides.
We use Apache DataFusion, which is the fastest single-node engine for querying Parquet files. We've tuned it for observability workloads.
You can query your data with SQL, no new query language to learn. This also means that your LLMs can easily help you (AI is very good at SQL), for example via our Logfire MCP server.
Everything's built on open standards (OTel, SQL, Parquet). Easy to extend, integrate, or migrate. No lock-in.
Get Started
For more information contact: [email protected]
Explore Logfire.
Explore our open source packages
