bl-mcp
Related Servers
Alternatives to bl-mcp
No user-submitted related servers found.
Related Servers
- AlicenseBqualityAmaintenanceEnables AI agents to drive a live Blender session for modeling, shading, lighting, rendering, and scene manipulation, returning rendered images for visual feedback.122MIT
- AlicenseAqualityBmaintenanceEnables AI agents to drive Blender 5.1+ for modeling, sculpting, mesh editing, scene management, and rendering through live Python execution.261MIT
- AlicenseCqualityAmaintenanceEnables AI agents to control Blender through 150+ tools across 24 modules, including modeling, materials, lighting, animation, rendering, mesh quality analysis, and visual feedback, plus expert prompts and goal-first workflow management.2091MIT
- FlicenseNot gradedqualityAmaintenanceEnables AI agents to control a live Blender session, inspect and modify scenes, manage PolyHaven assets, render images, and route prompts through configurable local or OpenAI-compatible LLM backends.-
- FlicenseNot gradedqualityAmaintenanceEnables AI agents to control a live Blender session, creating, modifying, and deleting objects, applying materials, rendering images, executing code, and pulling PolyHaven assets, while routing prompts through a provider-agnostic LLM backend that can be switched at runtime.-
- AlicenseNot gradedqualityBmaintenanceEnables MCP hosts and coding agents to model, inspect, measure, and deliver real Blender scenes through 15 tools and 217 catalog operations, returning actual rendered images after key edits. Works against a live Blender GUI via addon or a persistent headless worker, producing editable .blend files, GLB exports, previews, and acceptance records.1BSD 3-Clause
TDQS
Scored across 113 tools
Despite 113 tools, descriptions actively disambiguate overlapping operations by explicitly cross-referencing alternatives (subdivide vs subdivide_faces vs remesh; attach vs move_to_contact vs place_on vs ground; combine vs array vs mirror vs duplicate). The many face-editing tools (extrude_faces, inset_faces, delete_faces, sculpt, transform_region, etc.) share one `where` selector and similar framing, which creates some residual cognitive load, but each targets a distinct operation.
Names are thoroughly snake_case with predictable verb_noun or noun patterns (create_primitive, add_light, check_mesh, set_material, delete_faces). Minor deviations exist: many tools are noun-only (limb, blob, rock, terrain, loft, sweep, boolean) and some compound oddities (palette_uv, noise_displace), but casing and style stay consistent throughout.
113 tools is far beyond what any single agent can hold in working context, even for a domain as broad as full 3D modeling, and exceeds the rubric's 50+ 'extreme mismatch' band. The breadth of Blender (modeling, materials, rendering, rigging, level design, reference matching) partly justifies the volume, but the surface is heavy and hard to navigate as a set.
The surface covers an extraordinary lifecycle: primitives and mesh editing, transforms, materials/UV/baking, rendering and post, rigging and posing, export/import with checks, plus verification tools (assert_spec, check_game_ready, compare_view, diff_since) and safety nets (checkpoint, rollback, job_status). Almost no domain gap is left uncovered for a Blender automation agent.