Capture & CreateAI-made 3D

Meshy 7 Turns a Single Photo Into an Impressively Detailed 3D Model

We tested Meshy 7 with our own images. The detailed, fully rotatable results show how far image-to-3D generation has come.

ShareXFacebookLinkedIn
A textured 3D bust of Abraham Lincoln generated in Meshy 7 beside the portrait used as its single reference image
We generated this textured Abraham Lincoln bust in Meshy 7 from the single portrait shown at left.Image: Spatial Insider

Meshy 7 (opens in a new tab) is now available with an upgraded image-to-3D model that can turn a single picture into a textured, fully rotatable 3D object.

Meshy says the new model more accurately preserves a reference image’s proportions, part placement, and surface detail. Draft models take about 30 to 40 seconds to generate, according to the company. That lined up with our testing, where the results appeared well under a minute.

Meshy’s examples look impressive

The company’s launch examples include expressive faces, an intricate mechanical bird, and a carved jade medallion. A newer demonstration places the original images beside Meshy 7’s generated geometry and textured results, making the improvements easy to see.

Meshy calls the improvement alignment. In plain terms, the generated object should look more like the picture someone uploaded, with the right shapes in the right places instead of a loose 3D interpretation.

Then we tried it ourselves

The company’s examples looked good, but our own results were the real surprise.

We gave Meshy 7 a portrait of Abraham Lincoln. It returned a remarkably recognizable textured bust with modeled hair, clothing, ears, and even the back of his head. The original picture could not fully show many of those details.

A full rotation of the textured Abraham Lincoln model we generated from one portrait in Meshy 7.Video: Spatial Insider

The result is not merely a flat image wrapped around a simple shape. Rotate the model and Lincoln’s profile, hair, beard, collar, bow tie, and jacket continue around the entire bust. It is genuinely incredible to see this come from one picture.

For a different kind of complexity, we tried a rotary phone. Lincoln tested facial likeness, hair, and clothing, while the phone presented mechanical shapes, a circular dial, separate feet, and a thin coiled cord.

Meshy 7 reconstructed the phone as a complete object, including geometry hidden from the camera in the original image. The cord loops around the back, the handset rests above the body, and the circular dial openings remain recognizable as the model turns.

The generated phone remains recognizable through a full rotation, including its dial, feet, and coiled cord.Video: Spatial Insider

From one photo to games, AR, and 3D prints

Meshy can export models as GLB for the web, FBX for game engines, STL for 3D printing, and USDZ for Apple’s AR Quick Look, among other formats. Creators can take a generated model beyond the preview and bring it into the tools where they already work.

A game developer could quickly create a prop from concept art. A spatial designer could turn a product image into an object for an AR scene. A 3D artist could generate a base model and spend more time refining it instead of beginning from an empty canvas.

Meshy 7 gave us two impressively detailed objects we could immediately rotate, inspect, texture, export, and build on. Both already looked polished inside the viewer.

Image-to-3D generation is getting very good, very fast. Our results made it easy to imagine using Meshy 7 for real creative work today.

More from Capture & Create.

Browse all stories
A hand pinches an interactive 3D model beside the Spatial Insider website in Safari on Apple Vision Pro

Spatial Web · 4 min read

Safari Lets You Drag 3D Models Into Your Space on Vision Pro

A USDZ file and one HTML element are enough to make a web model interactive, stereoscopic, and draggable at real size.

A gardener beside a Gaussian splat preview in a flower-filled garden

Gaussian splats · 1 min read

Intangible’s Browser-Based 3D Studio Gains Gaussian Splat Support

The update lets creators place scans of real places beside other 3D objects and move a camera through the scene.

Front and back views of the Insta360 X6 camera

Insta360 X6 · 4 min read

Insta360 X6 Now Available With 8K50 and Spatial Capture

The $699.99 camera adds larger sensors, 120MP photos, longer 8K recording, and an easier way to create Gaussian splats.

Preparing the Spatial Insider studio