Skip to main content
En Croissant supports multiple chess engines for position analysis, evaluation, and game review. You can install engines from a curated list, add cloud-based engines, or configure your own local UCI-compatible engines.

Accessing Engine Settings

Navigate to the Engines page to view, install, and configure all your chess engines. Each engine displays:
  • Engine name and icon/image
  • ELO rating (if available)
  • Version information
  • File path (for local engines) or URL (for cloud engines)

Installing Engines

Configuring Engines

Select an engine from the list to configure its settings:

General Settings

  • Name: Display name shown in the interface
  • Version: Engine version identifier
  • ELO: Estimated strength rating
  • Image: Custom icon (click to change)

Search Settings

Configure how the engine searches for best moves:
  • Depth: Analyze to a specific depth (e.g., 20 moves)
  • Time: Analyze for a specific duration (in milliseconds)
  • Nodes: Analyze a specific number of positions
  • Infinite: Analyze until manually stopped

Advanced UCI Settings

En Croissant automatically detects and displays all UCI options supported by your engine:

Common UCI Options

MultiPV (Number of Lines)
  • Number of principal variations to display
  • Range: 1-500
  • Higher values show more alternative moves but slow analysis
Threads
  • Number of CPU threads to use
  • Default: Auto-detected based on your CPU
  • More threads = faster analysis
Hash
  • Memory allocated for the transposition table (in MB)
  • Default: 128 MB
  • Larger values improve analysis speed but use more RAM
Syzygy Tablebases
  • Path to endgame tablebase files
  • Provides perfect play in endgames
Required settings (MultiPV, Threads, Hash) are automatically configured when installing engines. Other options can be customized based on your needs.

Engine-Specific Options

Different engines may expose additional options:
  • Check options: Boolean toggles
  • Spin options: Numeric values with min/max ranges
  • Combo options: Dropdown selection from predefined values
  • String options: Text input fields
  • File options: File path selection (click to browse)

Managing Engines

Duplicate an Engine

Create a copy of an engine with different settings:
  1. Select the engine
  2. Click “Duplicate”
  3. Modify the name and settings as needed
This is useful for creating multiple configurations of the same engine (e.g., fast analysis vs. deep analysis).

Edit Engine JSON

For advanced users, click “Edit JSON” to directly modify the engine configuration:
{
  "type": "local",
  "id": "uuid",
  "name": "Stockfish 16",
  "version": "16",
  "path": "/path/to/stockfish",
  "elo": 3600,
  "settings": [
    { "name": "MultiPV", "value": 3 },
    { "name": "Threads", "value": 8 },
    { "name": "Hash", "value": 2048 }
  ]
}

Reset Engine Settings

Click “Reset” to restore all UCI options to their default values (while preserving required settings).

Remove an Engine

Click “Remove” to delete an engine from En Croissant. This does not delete the engine binary from your system.
If a local engine shows “(file missing)”, the executable has been moved or deleted. Update the path or reinstall the engine.

UCI Protocol Support

En Croissant fully supports the Universal Chess Interface (UCI) protocol, making it compatible with all standard UCI engines including:
  • Stockfish
  • Komodo
  • Leela Chess Zero (Lc0)
  • Dragon
  • And many others
The UCI implementation handles:
  • Engine initialization and communication
  • Position setup via FEN strings
  • Move sequences in UCI notation
  • Real-time evaluation updates
  • Multiple concurrent engine instances

Engine Storage

Engine configurations are stored in:
  • Location: engines/engines.json in your engines directory
  • Format: JSON with automatic schema validation
  • Persistence: Settings are saved automatically
To open your engines folder, click the folder icon next to the “Engines” title. Based on the default engines list, En Croissant recommends:
  • Stockfish: The strongest open-source engine (3600+ ELO)
  • Komodo Dragon: Strong commercial engine with unique evaluation
  • Leela Chess Zero: Neural network-based engine with human-like play
Engine availability and versions are automatically updated based on your platform.

Build docs developers (and LLMs) love