Langflow Observability & Monitoring with OpenTelemetry

Langflow 1.10+ now includes built-in OpenTelemetry support for real-time observability of visual AI workflows. Users can export flow and LLM traces to SigNoz by running the official Docker image with environment variables, enabling end-to-end tracing, alerts on latency and errors, and insights into components, model calls, prompts, and token usage.

What is Langflow Observability? Langflow observability gives you real-time visibility into your visual AI workflows by collecting traces using OpenTelemetry https://opentelemetry.io/ . Langflow 1.10+ ships with OpenTelemetry built in, so you don't build or install any instrumentation. This guide shows you how to export Langflow's flow and LLM traces to SigNoz by running the official Langflow Docker image with a few environment variables, giving you insight into components, model calls, prompts, and token usage. With full Langflow observability in SigNoz, you can trace every flow run end to end, correlate LLM spans carrying gen ai. attributes, set alerts on latency and errors, and continuously improve the reliability of your Langflow applications. Prerequisites - A SigNoz Cloud account https://signoz.io/teams/ with an active ingestion key or Self Hosted SigNoz instance https://signoz.io/docs/install/self-host/ - Docker installed and running on your system - Langflow 1.10 or later. Follow the Langflow installation guide https://docs.langflow.org/get-started-installation if you don't have it yet Monitor Langflow with OpenTelemetry Langflow's OpenTelemetry stack is bundled inside the Langflow image, so the integration is entirely configuration based. For more details, refer to the Langflow documentation https://docs.langflow.org/ . Step 1: Run Langflow with telemetry enabled Replace <your-ingestion-key and the region in the URL, then run the official Langflow image: docker run -d --name langflow -p 7860:7860 \ -e TRACELOOP API KEY=signoz \ -e TRACELOOP BASE URL=https://ingest.<region .signoz.cloud:443 \ -e TRACELOOP HEADERS="signoz-ingestion-key=<your-ingestion-key " \ -e TRACELOOP TRACE CONTENT=true \ langflowai/langflow:latest <region : Your SigNoz Cloud region https://signoz.io/docs/ingestion/signoz-cloud/overview/ endpoint us , eu , or in <your-ingestion-key : Your SigNoz ingestion key https://signoz.io/docs/ingestion/signoz-cloud/keys/ Langflow is now exporting traces to SigNoz. Keep the https:// scheme and :443 port. Langflow appends /v1/traces for you, so don't add it yourself. Most steps are identical. To adapt this guide, update the endpoint and remove the ingestion key header as shown in Cloud → Self-Hosted https://signoz.io/docs/ingestion/cloud-vs-self-hosted/ cloud-to-self-hosted . The following environment variables control the export: | Variable | Required | Value | What it does | |---|---|---|---| TRACELOOP API KEY | Yes | any non-empty string e.g. signoz | Switches Langflow's OTLP tracer on. It is not used to authenticate to SigNoz. | TRACELOOP BASE URL | Yes | https://ingest.<region .signoz.cloud:443 | Where Langflow sends OTLP data it posts to .../v1/traces . | TRACELOOP HEADERS | Yes | signoz-ingestion-key=<your-ingestion-key | Authenticates the export to SigNoz. | TRACELOOP TRACE CONTENT | No | true / false default true | Include prompt and response text on spans. Set false to keep prompt bodies out of telemetry while still capturing model and token data. | Prefer docker-compose? Set the same variables in your compose file or a .env file and start Langflow with docker compose up -d . The langflowai/langflow image reads the same TRACELOOP variables either way. Step 2: Generate some telemetry - Open Langflow at http://localhost:7860 - Build or open a flow that calls an LLM, and run it Each run produces spans, including LLM spans carrying gen ai. attributes such as the model name and token usage. An idle Langflow instance emits nothing, so a flow must actually run before data appears in SigNoz. View Langflow Traces in SigNoz Once configured, your Langflow application automatically emits traces for every flow run. Traces are available in SigNoz under the Traces tab: When you click on a trace in SigNoz, you'll see a detailed view of the trace, including all associated spans, along with their events and attributes. LLM spans carry gen ai. attributes model, token usage and, when TRACELOOP TRACE CONTENT=true , the prompt and response text. This integration sends Langflow's flow and LLM spans from its native tracer. Full HTTP-request and service-map instrumentation is a separate mechanism and is not covered here. Langflow Observability Dashboard You can also check out our custom Langflow dashboard https://signoz.io/docs/dashboards/dashboard-templates/langflow-dashboard/ which provides specialized visualizations for monitoring your Langflow usage. The dashboard includes pre-built charts specifically tailored for LLM usage, along with import instructions to get started quickly. Troubleshooting Langflow Observability Troubleshooting Langflow Observability troubleshooting-langflow-observability No traces in SigNoz - Confirm the flow actually ran. An idle Langflow instance emits nothing. - Verify TRACELOOP API KEY is set to a non-empty value. This switches the tracer on. - Check that the region in TRACELOOP BASE URL matches your SigNoz account. - OpenTelemetry batches data before sending, so wait 10-30 seconds after running a flow. Auth errors 401 / 403 Re-check the ingestion key in TRACELOOP HEADERS=signoz-ingestion-key=<key . It must be the exact key from your SigNoz Ingestion Settings, with no extra spaces or quotes. Endpoint issues Keep the https:// scheme and :443 port in TRACELOOP BASE URL . Langflow appends /v1/traces for you, so don't add it yourself. Upgrading Langflow Upgrading Langflow upgrading-langflow Nothing special is required. The OpenTelemetry stack is bundled inside the Langflow image, so it always matches the version you run. There is no separate agent or instrumentor to update: docker pull langflowai/langflow:latest or a specific tag recreate the container with the same env vars For reproducible deploys, pin a specific tag such as langflowai/langflow:1.10.1 instead of latest . Setup OpenTelemetry Collector Optional Setup OpenTelemetry Collector Optional setup-opentelemetry-collector-optional What is the OpenTelemetry Collector? Think of the OTel Collector as a middleman between your app and SigNoz. Instead of your application sending data directly to SigNoz, it sends everything to the Collector first, which then forwards it along. Why use it? Cleaning up data - Filter out noisy traces you don't care about, or remove sensitive info before it leaves your servers. Keeping your app lightweight - Let the Collector handle batching, retries, and compression instead of your application code. Adding context automatically - The Collector can tag your data with useful info like which Kubernetes pod or cloud region it came from. Future flexibility - Want to send data to multiple backends later? The Collector makes that easy without changing your app. See Switch from direct export to Collector https://signoz.io/docs/opentelemetry-collection-agents/opentelemetry-collector/switch-to-collector/ for step-by-step instructions to convert your setup. For more details, see Why use the OpenTelemetry Collector? https://signoz.io/docs/opentelemetry-collection-agents/opentelemetry-collector/why-to-use-collector/ and the Collector configuration guide https://signoz.io/docs/opentelemetry-collection-agents/opentelemetry-collector/configuration/ . Additional resources: - Set up alerts https://signoz.io/docs/alerts/ for high latency or error rates - Learn more about querying traces https://signoz.io/docs/userguide/traces/ - Explore log correlation https://signoz.io/docs/userguide/logs query builder/