20 open-source AI agent observability tools worth knowing in 2026, categorized
MaryamMiradi · x · 2026-11-06
A practical roundup of the open-source AI agent observability ecosystem for regulated workflows (medical, finance, legal). Key thesis: logging the final answer isn't observability — production agents need full-run visibility into model calls, tool calls, retrievals, errors, latency, and cost.
- Agent/LLM tracing: Langfuse, Arize Phoenix, Opik, AgentOps
- OpenTelemetry & instrumentation: OpenLIT, OpenLLMetry, Langtrace, OTEL Collector
- Logs & metrics: SigNoz, Grafana, Prometheus, Loki
- Cost & performance: Helicone, MLflow Tracing, W&B Weave, AgentMetrics
- Debugging & production: LangWatch, Pydantic Logfire, etc.
The author may benchmark Langfuse vs Phoenix vs Opik vs AgentOps vs OpenLIT vs Helicone next.
More from coding & agent
- Executor v2 launches: Zapier/Retool-like automation fully drivable by code and AI agents — threepointone · 2026-11-06
- Y Combinator-backed Reflex raises $14M Series A, its Python framework hits 29k GitHub stars — ycombinator · 2026-11-06
- One diagram maps all the agent protocols: MCP, A2A, UCP, AP2, A2UI, AG-UI — rseroter · 2026-11-06
- Sutra: open-source terminal coding agent, lowest cost per solved task — cneuralnetwork · 2026-11-06
- From Greasemonkey to now: AI lets you mod any app just by describing the change — msfeldstein · 2026-11-06
- AI agent handling an office move spams movers with hundreds of questions until they refuse to reply — flngr · 2026-11-06