Product
Pricing
Plans
Self-host OpenLIT free under Apache 2.0 for unlimited LLM tracing, evaluations, prompt management, and agent monitoring. OpenLIT Cloud is coming soon for teams that want a fully hosted option.
OSS
Self-host the open source Harness Engineering platform on your own infra with Helm or Docker.
Free
Apache 2.0. Unlimited self-hosted usage.
- Apache 2.0 licensed OpenLIT platform
- Full stack AI infra monitoring for GPUs, LLMs, MCP, vector DBs, and coding agents
- Core platform features and APIs for tracing, evaluation, prompts, and more
- Deployment docs and Helm chart
- OAuth sign-in
- Community support on GitHub
Cloud
Fully hosted OpenLIT for teams that prefer managed operations.
Coming soon
Join the waitlist for launch updates and pricing.
- Fully hosted OpenLIT platform
- Managed infrastructure and upgrades
- Built for teams that prefer zero ops
- Feature set and pricing shared at launch
What is included in free OpenLIT OSS
Everything below ships today in the open source, self-hosted OpenLIT platform.
Zero-code instrumentation
eBPF and SDK auto-instrumentation across Kubernetes, Docker, and Linux.
- Controller binary and deployment options
Coding agent observability
Session, cost, and code-impact tracking for coding agents.
- Session, cost, token, and repo activity tracking
Fleet Hub
OpenTelemetry collector monitoring and config via OpAMP.
- Single-cluster collector monitoring
Prompt Hub
Prompt versioning, retrieval, and AI-assisted improvement.
- Prompt versioning
- Dynamic variables and API retrieval
- UnlimitedPrompt fetching
- AI-assisted prompt improvement
Vault
Encrypted secret storage for LLM API keys.
- Key and tag based retrieval
- API authentication
- Encrypted storage (AES-256-GCM)
OpenGround
Side-by-side LLM comparison for latency, tokens, and cost.
- BYOK model comparison
Custom dashboards
ClickHouse-backed dashboards with charts, tables, and filters.
- Dashboard creation, widgets, and folders
- Import and export
Evaluations
Built-in evaluation types with rule-engine automation.
- Built-in evaluation types
- Programmatic and local or CI execution
- Custom eval prompts and rule-based selection
Guardrails
Prompt-injection, jailbreak, and sensitive-topic detection.
- Built-in guards in SDK (LLM and regex)
GPU monitoring
GPU utilization, memory, temperature, and power visibility.
- Collector and dashboards
Usage and limits
- Unlimited (self-hosted)Included usage
- Unlimited (self-managed)Access to historical data
- UnlimitedUsers
- UnlimitedProjects and environments
Support
- Community support (GitHub)
Security and compliance
- Sign in with Google, Azure AD, GitHub
| Feature | Included in OSS |
|---|---|
Zero-code instrumentation eBPF and SDK auto-instrumentation across Kubernetes, Docker, and Linux. | |
| Controller binary and deployment options | |
Coding agent observability Session, cost, and code-impact tracking for coding agents. | |
| Session, cost, token, and repo activity tracking | |
Fleet Hub OpenTelemetry collector monitoring and config via OpAMP. | |
| Single-cluster collector monitoring | |
Prompt Hub Prompt versioning, retrieval, and AI-assisted improvement. | |
| Prompt versioning | |
| Dynamic variables and API retrieval | |
| Prompt fetching | Unlimited |
| AI-assisted prompt improvement | |
Vault Encrypted secret storage for LLM API keys. | |
| Key and tag based retrieval | |
| API authentication | |
| Encrypted storage (AES-256-GCM) | |
OpenGround Side-by-side LLM comparison for latency, tokens, and cost. | |
| BYOK model comparison | |
Custom dashboards ClickHouse-backed dashboards with charts, tables, and filters. | |
| Dashboard creation, widgets, and folders | |
| Import and export | |
Evaluations Built-in evaluation types with rule-engine automation. | |
| Built-in evaluation types | |
| Programmatic and local or CI execution | |
| Custom eval prompts and rule-based selection | |
Guardrails Prompt-injection, jailbreak, and sensitive-topic detection. | |
| Built-in guards in SDK (LLM and regex) | |
GPU monitoring GPU utilization, memory, temperature, and power visibility. | |
| Collector and dashboards | |
Usage and limits | |
| Included usage | Unlimited (self-hosted) |
| Access to historical data | Unlimited (self-managed) |
| Users | Unlimited |
| Projects and environments | Unlimited |
Support | |
| Community support (GitHub) | |
Security and compliance | |
| Sign in with Google, Azure AD, GitHub | |
OpenLIT pricing FAQ
What is OpenLIT pricing?
OpenLIT OSS is free to self-host under Apache 2.0 with no license key and no per-trace fee. OpenLIT Cloud is coming soon as a fully hosted option. Cloud pricing will be shared at launch.
Is OpenLIT free and open source?
Yes. OpenLIT is an open source Harness Engineering platform. You can self-host it free forever under Apache 2.0. Deploy with Docker Compose or Kubernetes Helm from GitHub.
Can I self-host OpenLIT for LLM observability?
Yes. Self-host OpenLIT on your own infrastructure. The stack includes OpenLIT, ClickHouse, and an OpenTelemetry Collector. Usage on OSS is unlimited because you run it yourself.
Is OpenLIT Apache 2.0 licensed?
Yes. OpenLIT is Apache 2.0 licensed. You can inspect the code, self-host, modify, and contribute without a paid license for OSS.
What is OpenLIT Cloud?
OpenLIT Cloud is the fully hosted version of OpenLIT for teams that want managed operations instead of running their own stack. Feature and pricing details will be shared at launch. Join the waitlist for updates.
When will OpenLIT Cloud pricing be available?
Cloud is in development. Pricing details will be published at launch. Email [email protected] to join the waitlist.
Does self-hosted OpenLIT support OpenTelemetry?
Yes. OpenLIT is OpenTelemetry-native. Use OpenLIT SDKs or send OTLP to OpenLIT, and export traces to Grafana, Datadog, or any OTLP backend.
What is included in OpenLIT OSS?
OSS includes LLM tracing, coding agent observability, Prompt Hub, Vault, OpenGround, evaluations, guardrails, GPU monitoring, Fleet Hub, custom dashboards, OAuth, and community support on GitHub.
Questions? Email [email protected]
Get started
Ready to use the OpenLIT UI in production?
Self-host or connect your stack in minutes. Same harness UI for traces, dashboards, prompts, and evaluations.