Build a Bathtub Racer with GPT-6 Astra and V2Fun

Overview

Bathtub Racer No. 17 pairs a rounded, confident driver with an improvised racing machine: a blue enamel bathtub, exposed wheels and suspension, tubular frame, and an elevated rear engine. The design challenge is the relationship between them—the reclined torso, raised boots, steering wheel, and tub opening must read together.

The case uses a hybrid modeling approach. The technical notes assign visual analysis, decomposition, and assembly coordination to GPT-6 Astra. The component presentation identifies the seated driver as V2Fun-generated 3D and all four vehicle groups as Three.js procedural work. This division illustrates the broader GPT-6 Astra and V2Fun collaboration workflow.

Tools Used

• GPT-6 Astra: reference interpretation, component planning, and agent-side assembly coordination, as described in the case notes.
• V2Fun: generated character references and the textured seated-driver component.
• Three.js: procedural construction of the bathtub, chassis and steering, wheels and suspension, and rear engine.
• GLB: the supplied final asset format.

Step 1 — Establish the Modeling References

The input board presents a complete racer reference, an isolated driver image with the vehicle removed, and a front, side, and rear study. The complete reference establishes the vehicle silhouette and material palette. The isolated image makes the driver's reclining pose, gloves, helmet, clothing folds, and forward-facing boot soles easier to read.

The additional views are labeled AI-inferred. They provide design guidance for obscured areas, but do not independently verify the back or hidden mechanical structure. The single-image and multi-view input guide provides related guidance on preparing references and handling missing visual information.

Step 2 — Divide the Asset into Five Groups

The exploded presentation documents five component groups:

1. Seated driver: a continuous character with helmet, goggles, gloves, and raised boots.
2. Enamel bathtub: an open blue shell with an ivory rim, seat, and race number.
3. Chassis and steering: tubular frame, roll bar, steering wheel, and shaft.
4. Wheels and suspension: four matching wheels with hubs, tread, and coil springs.
5. Rear engine: finned cylinder banks, fuel tank, exhaust, and ignition wiring.

The spacing in this presentation is for displaying the parts. It does not establish detachable joints or a physical assembly system.

Step 3 — Generate the Driver and Build the Vehicle

V2Fun handles the driver as one character component. The process and final images show the rounded face, tan racing suit, striped helmet, goggles, and detailed boot soles. This is the case's documented use of image-to-3D model generation.

Three.js supplies the vehicle geometry, including the curved bathtub body and mechanical groups. The visible construction uses repeated wheel tread and springs, tubular supports, engine fins, and exhaust shields. The evidence supports this tool split without establishing a separate AI texturing pass or particular PBR map settings.

Step 4 — Assemble and Present the Model

The final view brings the driver into the tub with both boots raised over its front edge and the steering wheel positioned between the hands. The wheels surround the exposed chassis, while the engine and roll frame rise behind the driver.

The final presentation identifies its image as a render of the delivered GLB after reloading. It shows the completed static assembly from a slightly elevated front three-quarter angle. Hidden structures remain inferred.

Workflow

Complete reference → isolated driver and inferred view studies → five component groups → V2Fun driver generation and Three.js vehicle construction → assembly → GLB export → reloaded model presentation.

Prompt

English translation of the design brief; these are requested attributes rather than independently verified technical properties:

Create a rounded adult male racing driver with a confident smile, wearing a cream-and-brown striped vintage helmet, transparent goggles, a tan racing jumpsuit, leather gloves, lace-up work boots, and a small No. 17 chest patch. Recline him inside a blue enamel bathtub converted into a four-wheel racer. Both hands hold a small metal steering wheel, with knees bent high and both boot soles facing forward.

Give the bathtub a white rolled rim, a large white 17 on its blue side, chipped paint, light rust, and scuffs. Include four black rubber tires with recessed silver-gray hubs, exposed tubular chassis, front axle, steering links, copper-orange coil suspension, curved guards, two differently colored round headlights, and a small license plate. Behind the driver, place a high exposed engine with cooling fins, red and blue lines, a tubular roll cage, and a rearward exhaust with a perforated heat shield.

Combine humorous realism with lightly cartooned proportions and a collectible-model appearance. Request clear clothing folds and boot tread, refined PBR metal, leather, and painted surfaces. Show the complete driver and vehicle from a slightly elevated front three-quarter view against a white studio background, with soft lighting and contact shadows. Use a square composition with one standalone asset, no extra scenery, and no watermark.

Result

The final render retains the central visual identity: the smiling helmeted driver, raised boots, blue tub with white rim and No. 17 marking, dark treaded tires, copper-colored suspension, and exposed rear machinery. It demonstrates a static character-and-vehicle assembly; exact reference matching and material-map quality were not measured.

Supplied output metadata: bathtub-racer.glb, 52,156,048 bytes, MIME type model/gltf-binary. The case provides no generation-time or token-use measurements. For subsequent format decisions, see the GLB and other 3D export formats guide.