optuna/optuna-mcp

Overview

An MCP server that exposes Optuna's hyperparameter optimization APIs through the Model Context Protocol. It allows AI agents and MCP-compatible tools to query study results, create new studies, and interact with optimization trials.

Best for

Best for
Developers building MCP-compatible AI assistants that need hyperparameter optimization capabilities

Use cases

• Querying historical hyperparameter optimization studies from an AI agent
• Creating and managing Optuna studies via natural language commands
• Integrating Optuna into MCP-compatible IDEs or chat interfaces

How to use

Tools exposed

• create_study
• set_sampler
• get_all_study_names
• set_metric_names
• get_metric_names
• get_directions
• get_trials
• best_trial
• best_trials
• ask
• tell
• set_trial_user_attr
• get_trial_user_attrs
• plot_optimization_history
• plot_hypervolume_history
• plot_pareto_front
• plot_contour
• plot_parallel_coordinate
• plot_slice
• plot_param_importances

Tested with

Claude Desktop

Example client config

{\n  "mcpServers": {\n    "Optuna": {\n      "command": "/path/to/uvx",\n      "args": [\n        "optuna-mcp"\n      ]\n    }\n  }\n}

Notes

An MCP server that exposes Optuna’s hyperparameter optimization APIs through the Model Context Protocol. It allows AI agents and MCP-compatible tools to query study results, create new studies, and interact with optimization trials.

76 stars on GitHub. Last updated 2026-05-28. Licensed MIT.

Use cases

• Querying historical hyperparameter optimization studies from an AI agent
• Creating and managing Optuna studies via natural language commands
• Integrating Optuna into MCP-compatible IDEs or chat interfaces

Pros

• Open-source and Python-based, integrates seamlessly with existing Optuna workflows
• Provides a standardized interface for AI tools to access optimization data
• Lightweight server with minimal setup overhead

Cons

• Relatively new project with limited community adoption and documentation
• Depends on Optuna’s API, which may undergo breaking changes across versions
• Useful only for teams already using Optuna for hyperparameter tuning

Indexed from awesome-mcp-servers-punkpeye and enriched against its public facts.