~ same prompt, different LLM — every model gets the same brief in Blender, you judge the render ~
OpenAI / GPT 6 Astra / thinking:max has entered the arena NEW!
Render Window — hero
hero render
camera + framing set up by the AI contestant
3D Viewport — cabin-pond User Persp
LOADING 3D VIEW
$0.04Cost
38sDuration
5Tool calls
6Turns
Blender MCP 2026-07-14
Tokens: 30457/3095Vision: Votes W-L-T: 5-3-0DONE ✓
B-BENCH — TRANSCRIPT OpenAI / GPT 5.6 Luna · r1 · 2026-07-14
RUN LOG — cabin-pond · OpenAI / GPT 5.6 Luna · r1 · ok
SYSTEM PROMPTthe control — every model on this task got these instructions
SHOW FULL PROMPT · 432 chars
You are operating Blender through the provided tools. Build what the user's prompt asks for in the current scene. Work step by step and verify your own work. Before you finish, set up a camera and lighting to frame a single, well-composed presentation image of your result — position and aim the camera, and add or adjust lights so the subject reads clearly. When you are finished, reply with the single word DONE and no tool calls.
--TASK cabin-pond: A warm wooden cabin beside a pond, with a stone chimney, a small dock, and a few trees.
45:05>>get_scene_info({"user_prompt":"A warm wooden cabin beside a pond, with a stone chimney, a small dock, and a few trees."})
45:05<<{ "name": "Scene", "object_count": 2, "objects": [ { "name": "Cube", "type": "MESH", "location": [ 0.0, 0.0, 0.0 ] }, { "name": "Light", "type": "LIGHT", "location": [ 4.08, 1.01, 5.9 ] } ], "materials_count": 2 }
45:18>>execute_blender_code({"code":"import bpy, math\nfrom mathutils import Vector\n# clear\nbpy.ops.object.select_all(action='SELECT'); bpy.ops.object.delete(use_global=False)\nfor d in (bpy.data.materials, bpy.data.curves, bpy.data.cameras, bpy.data.lights):\n pass\n# materials helper\ndef mat(name, color, rough=0.7, metallic=0):\n m=bpy.data.materials.get(name) or bpy.data.materials.new(name)\n m.diffuse_color=(…)
45:18<<Error executing code: Communication error with Blender: Code execution error: bpy_struct: item.attr = val: enum "BLENDER_EEVEE_NEXT" not found in ('BLENDER_EEVEE', 'BLENDER_WORKBENCH', 'CYCLES')
45:22>>execute_blender_code({"code":"import bpy\nfrom mathutils import Vector\nsc=bpy.context.scene\nsc.render.engine='BLENDER_EEVEE'\n# ensure camera and lights if previous code got through\nif 'Presentation Camera' not in bpy.data.objects:\n bpy.ops.object.camera_add(location=(11,-16,9)); cam=bpy.context.object;cam.name='Presentation Camera';sc.camera=cam\nelse: cam=bpy.data.objects['Presentation Camera']; sc.camera=cam…)
45:22<<Error executing code: Communication error with Blender: Code execution error: Error: Cannot open file <repo>\harness\tmp\cabin_scene.blend@ for writing: No such file or directory
45:24>>execute_blender_code({"code":"import bpy\nfrom mathutils import Vector\nsc=bpy.context.scene; sc.render.engine='BLENDER_EEVEE'\ncam=bpy.data.objects.get('Presentation Camera')\nif not cam:\n bpy.ops.object.camera_add(location=(11,-16,9));cam=bpy.context.object;cam.name='Presentation Camera'\nsc.camera=cam;cam.rotation_euler=(Vector((-0.5,0.0,1.7))-cam.location).to_track_quat('-Z','Y').to_euler();cam.data.lens=48\nsc.r…)
45:25<<Code executed successfully: done 56
45:26>>get_viewport_screenshot({"max_size":800,"user_prompt":"A warm wooden cabin beside a pond, with a stone chimney, a small dock, and a few trees."})
45:28<<
45:28imgimage delivered — get_viewport_screenshot
45:31ai>DONE
45:31--status: ok · DONE declared
>
FULL CALL INSPECTOR
OPEN RAW JSONL ↗
0 SHOWN
LOADING COMPLETE TRANSCRIPT…
Votes cast on this bench: 000482 · 157 runs · 25 contestants
Best viewed with Netscape Navigator 4.07 at 800×600 · real runs, live ratings
Blender-Bench — which robot blends best?