Open Source
Free
- Self-hosted core under AGPL-3.0
Respan Gateway: up to 55% off tokens on Gemini, GPT, Claude and 1,000+ more models. One API key, 0% fee.
OpenObserve is an open-source observability platform for logs, metrics, traces, real user monitoring, and LLM applications, built by OpenObserve Inc. in Menlo Park. The core is written in Rust, stores telemetry as columnar Parquet with no index, and is licensed under AGPL-3.0, with a managed cloud and self-hosted editions.
LLM observability runs on OpenTelemetry. Traces capture the model, token counts, latency, and cost of each call, with integrations for the major providers, agent frameworks, and gateways like LiteLLM, Portkey, and OpenRouter. Online evaluations score live traces with an LLM judge or a remote scorer, and experiments run datasets through scorers offline, much like dedicated LLM evaluation tools, though both are enterprise features.
Pricing is per GB rather than per host or seat: Cloud charges $0.50 per GB ingested and $0.01 per GB queried with unlimited users, and the Self-Hosted Enterprise edition is free up to 50 GB a day. As a general observability platform, OpenObserve doesn't include an LLM gateway or prompt management, so routing and prompt versions stay in separate tools.
Core capabilities this platform advertises.
What this tool does well, and the limitations to keep in mind.
Pros
Cons
What's included in each plan, and how the tiers compare.
Free
Free
Up to 50 GB/day
$0.50/GB ingested + $0.01/GB queried
Per GB
Custom
Contact sales for a quote
Platform and SRE teams that want infrastructure and LLM telemetry in one open-source observability stack
OpenObserve covers infrastructure and LLM telemetry in one place. Respan is built for the LLM layer specifically: the AI router with observability and automated evals built in, so every model call is routed, traced, priced, and scored together, and evals are included from the free tier.
Respan is the AI router with observability and automated evals built in, so every LLM call is routed, traced, priced, and scored in one place. Evals are included from the free plan. Start free.
Top companies in Observability, Prompts & Evals you can use instead of OpenObserve.
Respan
Nested traces for every LLM call, tool run, and agent turn
LangSmith
Native tracing for LangChain and LangGraph
MLflow
OpenTelemetry-based tracing for LLMs and agents
Weights & Biases
ML experiment tracking
Arize AI
ML observability with LLM support
Langfuse
Tracing for LLM calls, tool runs, and agent steps
Helicone
Traceloop
OpenTelemetry
Datadog LLM
LLM monitoring within Datadog platform
Braintrust
Experiments comparing prompts and models
HoneyHive
Prompt management
Promptfoo
Patronus AI
Automated LLM evaluation platform
Phoenix
OpenTelemetry-based LLM and agent tracing
Humanloop
Portkey
DeepEval
Ragas
RAG-specific evaluation framework
Sentry
LangWatch
Multi-turn agent simulation testing
Galileo AI
LLM output quality evaluation
PromptLayer
Confident AI
DeepEval open-source evaluation framework
Maxim AI
Distributed tracing for LLM and agent apps
Opik
Agenta
Future AGI
Multimodal evaluation (text, image, audio, video)
Lunary
Parea AI
Athina AI
Ashr
Multi-modal synthetic testing
Sentrial
Agent failure root cause analysis
Chamber
ML infrastructure automation
Moda
Hallucination detection
Fiddler AI
Agentic observability across multi-step agent workflows
Laminar
Agent tracing with a readable transcript view
AgentOps
Session replay with time-travel debugging
Prefactor
Traces and spans for every agent run, with cost and data risk per step
Side-by-side comparisons with other tools in this category.
Companies from adjacent layers in the AI stack that work well with OpenObserve.