Ludo AI MCP

Ludo AI's MCP server puts the studio's game-asset pipeline inside any MCP-capable assistant. Instead of switching to a web app, you describe the asset and the agent submits the job, polls it and hands back the files.

The tool surface is unusually broad for a creative-tools server. Image tools produce sprites, icons, UI elements, textures and backgrounds, with background removal and upscaling. 3D tools turn 2D images into GLB models with PBR textures, auto-rig them with engine-ready joint naming, and drive skeletal animation from text or retarget animation presets. Directional-sprite tools render a 3D model from up to eight aligned angles in styles from pixel art to painted. Animation tools build spritesheets from static sprites (4-64 frames), transfer motion from video or presets, and repair or re-render sheets. Short videos, sound effects, music, character voices and text-to-speech round out the set.

Generation runs on an asynchronous job queue: a tool returns a job id immediately and getApiJob returns the result when it finishes; queued jobs can be cancelled for a credit refund. Several models are selectable per task, from a cost-effective pixel-art model to a more detailed one, with per-second credit pricing documented in the repo. A searchDocs tool lets the agent read Ludo's own guidance on animation modes, margins and known limitations before spending credits, and account tools report your balance first.

You need a Ludo.ai account and API key; usage is billed in credits. Best for indie developers and jam teams who want an assistant to produce and iterate on asset batches directly in their engine workflow.