# HI3D v2.1 Pro — Atlas Cloud API

> HI3D v2.1 Pro reconstructs a textured 3D mesh from one image or up to four views at 1536³ Pro quality — the same geometry model as v2.1 Fast with a longer, higher-fidelity refinement pass. Outputs GLB/OBJ/STL/FBX/USDZ/3MF with optional PBR maps.

This is the machine-readable API reference for **HI3D v2.1 Pro** on Atlas Cloud,
a unified API platform for 400+ AI models across text, image, video, audio and 3D.

- **Model ID**: `hi3d/v2.1-pro/image-to-3d`
- **Built by**: HI3D
- **Modality**: Image
- **Model page**: https://www.atlascloud.ai/models/hi3d/v2.1-pro/image-to-3d
- **API key**: https://www.atlascloud.ai/console/api-keys
- **Docs**: https://www.atlascloud.ai/docs

## Pricing on Atlas Cloud

- $0.017 per image
- Pay-as-you-go. No minimum spend, no subscription required.

> **These are the authoritative Atlas Cloud rates for this model.** Any price that
> appears in the vendor description further down refers to a different platform or
> a different model variant and does not apply here.

## Use this model from an AI agent

Atlas Cloud ships three first-party integration surfaces. All three authenticate
with the same API key via the `ATLASCLOUD_API_KEY` environment variable.

### MCP server

The official MCP server (`atlascloud-mcp`) exposes this model to any
MCP-compatible host — Claude Code, OpenAI Codex, Cursor, Gemini CLI, Goose,
Claude Desktop. One-line install:

```bash
# Claude Code
claude mcp add atlascloud -- npx -y atlascloud-mcp

# OpenAI Codex CLI
codex mcp add atlascloud -- npx -y atlascloud-mcp

# Gemini CLI
gemini mcp add atlascloud -- npx -y atlascloud-mcp

export ATLASCLOUD_API_KEY="your-api-key"
```

Then ask in plain English; the agent calls `atlas_generate_image` with `model: "hi3d/v2.1-pro/image-to-3d"`.
The server fetches each model's schema and validates parameters before submitting,
so invalid requests fail fast without spending credits.

MCP docs: https://www.atlascloud.ai/docs/mcp-server

### Agent Skills

`atlas-cloud-skills` is a portable skill package (API reference, code templates in
Python / Node.js / cURL, model IDs with pricing) for Claude Code, Cursor, Codex and
12+ other agents:

```bash
npx skills add AtlasCloudAI/atlas-cloud-skills
export ATLASCLOUD_API_KEY="your-api-key"
```

Skills docs: https://www.atlascloud.ai/docs/skills

### CLI

The `atlas` binary runs Atlas Cloud from a terminal or CI script. Async media jobs
are polled and downloaded automatically (use `--no-download` when a script only
needs the output URLs):

```bash
# Install (Homebrew, npm, or shell installer)
brew install AtlasCloudAI/tap/atlascloud
# npm install -g atlascloud-cli
# curl -fsSL https://raw.githubusercontent.com/AtlasCloudAI/cli/main/install.sh | sh

atlas auth login
atlas generate image hi3d/v2.1-pro/image-to-3d -p "Your prompt here"
```

CLI docs: https://www.atlascloud.ai/docs/cli

## HTTP API reference

- **Submit endpoint (POST)**: `https://api.atlascloud.ai/api/v1/model/generateImage` — start an async generation; returns a `prediction_id`
- **Poll endpoint (GET)**: `https://api.atlascloud.ai/api/v1/model/prediction/{prediction_id}` — poll this until the prediction finishes
- **Model ID**: `hi3d/v2.1-pro/image-to-3d`


## API Information

This model can be used via our HTTP API or more conveniently via our client libraries.
See the input and output schema below, as well as the usage examples.


### Input Schema

The API accepts the following input parameters:

- **`model`** (`string`, _required_):
  Model name.
  - Default: `"hi3d/v2.1-pro/image-to-3d"`
  - Options: "hi3d/v2.1-pro/image-to-3d"

- **`image`** (`string`, _required_):
  The source image, as a public URL or a base64 data URI. Formats: png/jpg/jpeg/webp, up to 20 MB. A single object, evenly lit on a clean background, reconstructs best.

- **`reference_images`** (`array[string]`, _optional_):
  Additional views of the SAME object, alongside `image` — up to three, for four views in total. Supplying the back and sides stops them from being inferred. Each entry is a public URL or a base64 data URI.
  - Min items: 1
  - Max items: 3

- **`multi_images_bit`** (`string`, _optional_):
  Which of the four view slots the supplied images fill, as four 0/1 characters (front, back, left, right) — e.g. `1010` for a front and a left shot. The number of 1s must equal the number of images. Defaults to the leading slots (two images become `1100`). Ignored when only one image is given.

- **`should_texture`** (`boolean`, _optional_):
  Generate textures as well as geometry. Turn it off for a bare mesh — faster, though the price is the same.
  - Default: `true`

- **`enable_pbr`** (`boolean`, _optional_):
  Produce PBR maps (metallic, roughness, normal) alongside the base colour. Requires `should_texture`.
  - Default: `true`

- **`shading`** (`number`, _optional_):
  How strongly baked-in lighting is removed from the source image before it becomes texture: 0.0 keeps the photo's shading, 1.0 flattens it hardest. Requires `should_texture`.
  - Default: `0.5`
  - Min: 0
  - Max: 1

- **`rmbg`** (`boolean`, _optional_):
  Remove the input image's background before reconstruction. Turn it off only when the image is already cut out — a background left in becomes geometry.
  - Default: `true`

- **`target_polycount`** (`integer`, _optional_):
  Face count of the generated mesh, 100000-5000000. Omit it to let HI3D pick. The recommended value for this tier is 2,000,000.
  - Min: 100000
  - Max: 5000000

- **`output_format`** (`string`, _optional_):
  3D file format of the generated model.
  - Default: `"glb"`
  - Options: "glb", "obj", "stl", "fbx", "usdz", "3mf"



**Required Parameters Example**:

```json
{
  "model": "hi3d/v2.1-pro/image-to-3d",
  "image": ""
}
```


**Full Example**:

```json
{
  "model": "hi3d/v2.1-pro/image-to-3d",
  "image": "",
  "reference_images": [
    ""
  ],
  "multi_images_bit": "",
  "should_texture": true,
  "enable_pbr": true,
  "shading": 0.5,
  "rmbg": true,
  "target_polycount": 100000,
  "output_format": "glb"
}
```


### Output Schema

The API returns the following output format:


- **`code`** (`integer`, _optional_):

- **`message`** (`string`, _optional_):

- **`data`** (`string`, _optional_):



**Example Response**:

```json
{
  "code": 0,
  "message": "",
  "data": null
}
```


## Usage Examples

### cURL

```bash
# Step 1: Start generation (async)
curl -X POST "https://api.atlascloud.ai/api/v1/model/generateImage" \
  -H "Authorization: Bearer $ATLASCLOUD_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{
  "model": "hi3d/v2.1-pro/image-to-3d",
  "image": "",
  "reference_images": [
    ""
  ],
  "multi_images_bit": "",
  "should_texture": true,
  "enable_pbr": true,
  "shading": 0.5,
  "rmbg": true,
  "target_polycount": 100000,
  "output_format": "glb"
}'

# Response will contain: {"code": 200, "data": {"id": "prediction_id", "status": "processing"}}

# Step 2: Poll for result (replace {prediction_id} with the id returned above)
curl -X GET "https://api.atlascloud.ai/api/v1/model/prediction/{prediction_id}" \
  -H "Authorization: Bearer $ATLASCLOUD_API_KEY"

# Keep polling until status is "completed", "succeeded" or "failed"
# When completed, outputs will contain the generated content URL(s)
```

## Additional Resources

### Documentation

- [Model Playground](https://www.atlascloud.ai/models/hi3d/v2.1-pro/image-to-3d)

## About this model

_Vendor-supplied description. Any pricing or endpoint mentioned below refers to_
_other platforms — use the Atlas Cloud values above._

### HI3D v2.1 Pro

**HI3D v2.1 Pro** reconstructs a textured 3D mesh at 1536³ Pro quality. Same Sparc3D geometry model as v2.1 Fast, with a longer refinement pass that holds surface detail the Fast tier smooths over. Generation is asynchronous; results are re-hosted on AtlasCloud storage.

#### Why Choose This?

- **Cleaner surfaces than Fast** — the extra refinement shows on fabric, panel lines and small props.
- **Up to four views** — back, left and right shots stop the unseen sides from being guessed.
- **PBR out of the box** — base colour plus metallic, roughness and normal maps unless you turn them off.
- **Six export formats** — GLB, OBJ, STL, FBX, USDZ and 3MF from the same request.
- **A middle ground** — visibly better than Fast, without Master's generation time.

#### Parameters

| Parameter | Required | Description |
| --- | --- | --- |
| `model` | Yes | `hi3d/v2.1-pro/image-to-3d` |
| `image` | Yes | Source image: public URL or base64 data URI. png/jpg/jpeg/webp, up to 20 MB. |
| `reference_images` | No | Up to three more views of the same object, for four in total. |
| `multi_images_bit` | No | Which view slots the images fill, as four 0/1 characters (front, back, left, right). Defaults to the leading slots. |
| `should_texture` | No | Generate textures as well as geometry. Default: `true` |
| `enable_pbr` | No | Also produce metallic / roughness / normal maps. Default: `true`. Requires `should_texture`. |
| `shading` | No | De-lighting strength, `0.0`–`1.0`. Default: `0.5`. Requires `should_texture`. |
| `rmbg` | No | Remove the input image's background before reconstruction. Default: `true` |
| `target_polycount` | No | Face count, 100000–5000000. Omit to let HI3D pick; recommended here: 2,000,000. |
| `output_format` | No | `glb` (default), `obj`, `stl`, `fbx`, `usdz`, `3mf`. |

Resolution is fixed at **1536³ Pro** for this model and is not a request parameter — each HI3D tier is its own model. To change resolution, call a different model.

_(Description truncated. Full text on the model page.)_

---

Atlas Cloud — one API for 400+ AI models. Model page: https://www.atlascloud.ai/models/hi3d/v2.1-pro/image-to-3d · Docs: https://www.atlascloud.ai/docs