Comet ML

Comet ML MCP Connector for Claude

A+

Manage machine learning experiments via Comet — track model metrics, audit project workspaces, and inspect ML run parameters directly from any AI agent.

6 tools Official Updated Jun 28, 2026 Official Vinkius Partner

Connect your Comet ML account to any AI agent and take full control of your machine learning lifecycle through natural conversation.

What you can do

  • Experiment Tracking — List and audit machine learning runs to inspect performance metadata, tags, and live execution statuses
  • Numeric Metric Auditing — Retrieve high-precision numeric endpoints mapping metrics generated dynamically during your training loops
  • Parameter Inspection — Extract explicit ML properties like learning rates and configurations logged to specific experiment keys
  • Project & Workspace Navigation — Navigate through organizational namespaces and identify exactly where your ML research resides
  • Run Metadata Analysis — Discovered disconnected physical limits parsing explicit run structures, timing, and structural configurations

How it works

  1. Subscribe to this server
  2. Enter your Comet ML API Key (found in Account Settings > API Keys)
  3. Start auditing your ML experiments from Claude, Cursor, or any MCP-compatible client

Who is this for?

  • Data Scientists — track and compare model metrics across experiments without leaving the research flow
  • ML Engineers — audit experiment parameters and verify training configurations using natural language
  • AI Researchers — navigate through multiple workspaces and projects to organize complex ML trials
  • MLOps Teams — monitor active model evaluations and verify experiment completion statuses in real-time
mlopsexperiment-trackingmodel-evaluationllm-monitoringmodel-lifecycledata-science

6 tools expose this connector's capabilities to your AI agent.

list_workspaces

Identify bounded routing spaces inside the Headless Comet ML limits

list_projects

Perform structural extraction matching target Projects inside Comet

list_experiments

Discover explicit routing arrays structuring specific logged experiment limits

get_experiment

Retrieve explicit Cloud logging tracing explicit Payload IDs

get_experiment_metrics

Execute static mapping targeting exactly defined numeric bounds natively

get_experiment_params

Inspect internal properties detailing API taxonomy types

See how to talk to your AI agent using Comet ML.

List all projects in workspace 'research-team'

I found 4 projects in 'research-team': 'NLP-LLM-v2', 'Computer-Vision-Edge', 'Tabular-AutoML', and 'Staging-Tests'. Which one would you like to explore?

Get current metrics for experiment 'exp_abc123'

Retrieving metrics for 'exp_abc123'... Current Accuracy: 0.945, Loss: 0.12, Epoch: 45. The run is still active and performance is trending upwards.

What hyperparameters were used in experiment 'exp_789'?

Experiment 'exp_789' used: learning_rate: 0.001, batch_size: 32, optimizer: 'adam', and model_architecture: 'resnet50'. I have the full list of params if you need them.

Yes. Use the 'get_experiment_metrics' tool with the experiment key. The agent will pull the latest numeric logged endpoints, allowing you to monitor loss, accuracy, and other custom metrics as they are generated.

Related Connectors