← All models

Hi3D

3D Available Full comparison ↗

hitem3d/hi3d · by Hitem3D · feed-forward-3d-reconstruction

Pricing — 2 offering(s)

fal

View on fal ↗

Image-to-3D (v2.1, 1536 fast)

$0.020 / credit Current 2026-09-02 → present

Image-to-3D (v2.1, 1536 pro)

$0.020 / credit Current 2026-09-02 → present

Hi3D (Hitem3D)

Image-to-3D (v2.1, 1536 pro)

$0.020 / credit Current 2026-09-02 → present

Image-to-3D (v3.0, 2048 quality)

$0.020 / credit Current 2026-09-02 → present

Showing the active price and any recorded history. Full pricing history is available via the paid API — see API docs.

See how Hi3D fits into a cost-aware routing setup

Capability profile

geometry quality strong

prompt alignment moderate

texture quality moderate

topology control weak

rigging and animation weak

input flexibility weak

generation speed moderate

Operator guidance

Choose Hi3D when you want the highest hosted output resolution from a single image and don't need text input, rigging, or topology control. Call the first-party API (docs.hi3d.ai) for the v3.0 2048³ line; fal carries v1.5-v2.1 at the same $0.02/credit rate with no signup step. For text-to-3D, multi-view, or a full asset pipeline (rig / remesh / segmentation), Tripo or Meshy fit better; for open weights, Hunyuan3D or TRELLIS.

Use cases

High-resolution image-to-3D where output detail outweighs cost
Product / prop reconstruction from a single reference photo
Portrait-to-3D (dedicated portrait-optimised variants)

Limitations

Image-to-3D only — no text or multi-view conditioning
No rigging, remesh, segmentation, or standalone re-texture operations
Architecture undisclosed; quality claims are vendor / third-party-comparison based
Not open-weights
3D Arena standing not independently re-verified this pass (see benchmark_notes)