Zetta N2 — our newer neural model, one image in
Zetta N2
A second neural engine with a deliberately narrow surface: exactly one image, textured output. Fewer knobs, a different reconstruction, and available on paid plans.
What it does
Single-image reconstruction
One picture in, one textured mesh out. No view slots and no mode switches — the model has one job and does it.
Text to 3D, still
Prompts work because the studio writes a reference image first and submits that. You type a description; N2 receives a picture.
A second opinion
A different engine reconstructs the same photograph differently. When N1 misreads a subject, N2 is worth trying before you change the input.
Textured by default
Every N2 job returns the model’s own textured output. There is no untextured mode to choose or to pay for.
How it works
- Supply one imageA photograph, a render, or a prompt the studio converts into one. Background removal usually improves the result.
- Select Zetta N2Available on paid plans. The studio refuses the combination up front if you ask N2 for something it cannot do.
- GenerateN2 runs on a tighter capacity budget than N1, so a busy queue means waiting rather than failing — the job is accepted and held, not rejected.
- RefineThe result is an ordinary library model — every post-processing tool applies, whichever engine made it.
What people use it for
- A single strong reference photograph where you want the best single-image result available
- Comparing two reconstructions of the same input before committing to cleanup
- Prompt-driven generation where you do not need low-poly or untextured modes
- Subjects N1 reads poorly — a second engine is often a faster fix than a better photograph
What it cannot do
- One image only. No multi-view input, so the far side is always inferred
- No low-poly base — reduce afterwards with retopology instead
- No untextured geometry-only mode
- No sketch mode
- A prompt with no image cannot be submitted directly; the studio always supplies a picture
- Paid plans only
- One job in flight at a time by default, so throughput is lower than N1
The details
- Inputs
- Exactly one image
- View slots
- None
- Bases
- Textured only
- Modes
- None — the model’s default output
- Availability
- Paid plans
- Concurrency
- One task at a time by default