OpenViSTAI · Agents · Blockchain · Crypto
Submit a project
Menu
AI Evaluation & Observability

Project Monocle

Project Monocle is an open-source developer tool that instruments generative AI application code and captures telemetry traces for monitoring and observability. It provides TypeScript and Python instrumentation, an OpenTelemetry-compatible span and trace format, and a metamodel for tracing GenAI components.

Explore the practical guide ↓

Independent OpenViST project profile, based on linked sources.

Project Monocle — OpenViST project profile in AI Evaluation & Observability.
OpenViST project illustration. See the linked sources for product details.
Understand the project

Practical overview

When it fits

Use Monocle when you need code-level tracing for a GenAI application and want to capture telemetry without implementing a custom instrumentation layer. Its quickstart is especially relevant when you can run a Python application through Monocle or need to add instrumentation because you do not control its runtime.

Who it is for

  • Application developers who want to trace GenAI code across environments.
  • Platform engineers instrumenting production applications through wrapping rather than asking application developers to recode GenAI components.
  • GenAI component providers adding observability features to their products.
  • Enterprises consuming GenAI application traces in an existing open-source observability stack.

Documented uses

  • Instrument a GenAI application to capture execution traces.
  • Trace activity across application code, model inference, and data retrieval.
  • Save captured traces locally as JSON or export telemetry to local or cloud storage.
  • Provide a consistent tracing format across GenAI component providers.
Before you choose

Requirements & limits

The project documentation describes Monocle as fully open source and community driven.

  • The quickstart documents TypeScript and Python instrumentation.
  • Python tracing has separate paths depending on whether you control how the application is run.
  • Trace output is written to a local JSON file by default; the project also describes export to local and cloud storage.
  • Monocle uses an OpenTelemetry-compatible format and supplies its own implementation and package.

Check current requirements in the official documentation. Source-based guidance does not mean OpenViST has run or independently tested this project.

Source-based practical guide

Capture traces from a Python application

Use Monocle to generate traces from a Python application and locate the default JSON output.

Before you start

  • A Python GenAI application.
  • For the no-instrumentation route, control over how the application is run; otherwise, the documented instrumentation route is for an application whose runtime you do not control.

Steps

  1. Install the Monocle package with the command below.

    Documented command
    pip install monocle_apptrace

  2. If you control the application launch, run it with the command below.

    Documented command
    python -m monocle_apptrace your-app-name.py

  3. If you do not control the runtime, import setup_monocle_telemetry from monocle_apptrace.instrumentation.common.instrumentor and call it in the application's main() function with a workflow name.

  4. Look for the generated monocle_trace___.json file in the directory where the application runs.

Expected outcome

Monocle writes traces to a JSON file by default in the application's execution directory.

Guide limitations

  • The quickstart presents the launch-through-Monocle and in-code instrumentation paths as alternatives based on runtime control.
  • The documented default output is a local JSON file; this guide does not cover configuring another export destination.
Documentation 1 ↗
COMPARE OPTIONS

Compare options for Model inference and serving

These projects document a common task and share a project role. This is a source-based comparison, not an execution benchmark or a guarantee of interchangeability.

ProjectDocumented purposePlatformAccess and requirements
Project MonocleInstrument a GenAI application to capture execution traces.The project documentation describes Monocle as fully open source and community driven.
FMBenchCompare inference latency and transactions per minute across model-serving configurations.CLI, CloudFMBench is distributed as a Python package and used from the command line. The project is published under the MIT-0 license.
Evidence and transparency

Sources & review

monocle2ai.github.io ↗Guide documentation ↗
Source record details
Source availability
Active
Sources checked
12 October 2026
Original website title
Monocle | Project Monocle · Official homepage title · retained in its source language
How we review sourcesSuggest a correction