# Late-night laundromat built in Blender

> A rainy-night coin laundromat rendered in Blender: rows of washers and dryers, a folding table and cart, and fluorescent tubes. Claude Code with Claude Opus 5.5 built it through Blender MCP from a ChatGPT reference image, with no manual Blender work.

- Kind: Video
- Page: https://meetreference.com/v/late-night-laundromat-3d-scene-blender/
- Ready-to-use agent handoff: https://meetreference.com/v/late-night-laundromat-3d-scene-blender/agent.txt
- Exact prompt as plain text: https://meetreference.com/v/late-night-laundromat-3d-scene-blender/prompt.txt
- Authorship: AI-made, per the source
- Made with: Claude Code
- Model: Claude Opus 5.5
- Built with: Blender
- Services: ChatGPT
- Skills or plugins: Blender MCP
- Style: 3D & WebGL; 3d, blender, photoreal, night scene, interior, procedural materials, japanese
- Length and format: 10 sec, 16:9 (1920x1080 source)
- Shared by @hayashimon1 in X, Oct 10, 2026: https://x.com/hayashimon1/status/2108913484805406859
- Prompt fidelity: Partial. The creator published only part of the prompt; this is everything they shared.

## Before you start

Guidance from Reference, separate from the creator's instructions.

1. Read the reference and inspect the project, installed libraries, and available agent skills before writing code. Reuse suitable tools already present.
2. Prefer the creator-reported libraries and skills when they fit the task. Explain what useful additions help with, and install missing project dependencies or relevant skills from official sources when the environment and the user's existing permissions allow it. Respect the project's package manager and setup rules; do not ask again for routine setup already authorized.
3. Distinguish tools used by the creator from your own recommendations. A listed tool is not automatically required, and an empty tools or requirements list means the setup is unknown, not that nothing is needed.
4. Ask concise questions when missing inputs or choices materially affect the result: paid services, credentials, the user's assets, or a different production approach. Use existing answers and permissions; continue independent work while waiting.
5. If a useful library, skill, or service cannot be installed or used, explain an available alternative and its expected effect on quality, features, or effort. Get the user's choice before a material compromise; otherwise state the fallback and continue. Never silently simplify the result or claim to have used a tool you did not use.

Requirements reported by the source: Blender with Blender MCP connected to Claude Code; A reference image of the look you want.

## Recreating this video

These are the creator's published instructions, unedited.

To adapt it: Replace "深夜のコインランドリー" (a late-night laundromat) with your own place, and hand Claude your own reference image.

## Prompts

### 1. Prompt, with the reference image

```text
このイメージで深夜のコインランドリーをBlenderで作成
```

### 2. Follow-up 1

```text
不自然なところが多い
```

### 3. Follow-up 2

```text
看板の余白が多い
```

### 4. Follow-up 3

```text
床をもっと参考画像に
```

### 5. Final request

```text
レンダリングしてmp4にして
```

## What it looks like

Written by Reference from the video frames, not by the creator.

A photoreal, slightly grimy Japanese coin laundromat at night, lit by fluorescent tubes, seen from a low, wide camera.

- Color: Cool grey-green walls and floor, brushed steel machines, blue-black windows, white fluorescent light, with small accents of red digital readouts, a blue bucket and pink laundry.
- Type: Only in-scene print: Japanese notices on the wall, a price sign and stickers on the machines, and numbered labels on the dryers.
- Framing: 16:9 one-point perspective down the room: stacked dryers and washers on the right, glass doors and a slatted bench on the left, a folding table and wire laundry cart in the middle, and a door at the far wall.
- Structure: One continuous shot that moves from the doorway toward the table and cart, ending close on the cart and the dryer stack.

Storyboard, nine frames spread across the video: https://meetreference.com/media/late-night-laundromat-3d-scene-blender/storyboard.webp

- Frame 1 at 0.6s: https://meetreference.com/media/late-night-laundromat-3d-scene-blender/f1.webp
- Frame 2 at 1.7s: https://meetreference.com/media/late-night-laundromat-3d-scene-blender/f2.webp
- Frame 3 at 2.8s: https://meetreference.com/media/late-night-laundromat-3d-scene-blender/f3.webp
- Frame 4 at 3.9s: https://meetreference.com/media/late-night-laundromat-3d-scene-blender/f4.webp
- Frame 5 at 5.0s: https://meetreference.com/media/late-night-laundromat-3d-scene-blender/f5.webp
- Frame 6 at 6.1s: https://meetreference.com/media/late-night-laundromat-3d-scene-blender/f6.webp
- Frame 7 at 7.2s: https://meetreference.com/media/late-night-laundromat-3d-scene-blender/f7.webp
- Frame 8 at 8.3s: https://meetreference.com/media/late-night-laundromat-3d-scene-blender/f8.webp
- Frame 9 at 9.4s: https://meetreference.com/media/late-night-laundromat-3d-scene-blender/f9.webp

## How it was made, per the creator

- One-shot: no
- Notes: ChatGPT made a 2D image of the finished look, which went to Claude with the first prompt. The scene went from a box blockout to models (boolean holes in the drums, cloth simulation for the laundry) to materials. Every material is procedural, with one node group adding wear, dirt, dust and fingerprints. Signs were designed in 2D inside Blender and rendered as textures. After the quoted feedback rounds, the creator asked for dryer spin, light flicker, rain and a slow camera push before the render.

## Links

- Watch: https://x.com/hayashimon1/status/2108913484805406859
- Original post: https://x.com/hayashimon1/status/2108913484805406859
- Step-by-step thread: https://x.com/hayashimon1/status/2108913486575411350