# AI3D API v2 model: sana-sprint

Generate a text-to-image with Sana Sprint through fal.ai.

- AI3D model ID: `sana-sprint`
- Task: `text_to_image`
- AI3D provider adapter: `fal`
- Source image: not used; provide a prompt.
- Availability: Check `GET https://api.ai3d.dev/v2/models`; deployment configuration and key model allowlists can affect access.
- Shared implementation guide: <https://api.ai3d.dev/v2/llms.txt>
- AI3D API reference: <https://api.ai3d.dev/v2/openapi.json>

## Request

Use AI3D's v2 job API. Do not call the provider directly or put provider credentials in the client.

- Method and URL: `POST https://api.ai3d.dev/v2/jobs`
- Authentication: `Authorization: Bearer <AI3D_API_KEY>` with write permission and this model allowed on the key.
- Content type: `application/json`
- `image_url` and `video_url` are not used.
- `input` is required and limited to 16 KiB. Supported fields:
  - `prompt` (required, 1–4,000 characters)
  - `negative_prompt` (optional string, default `""`)
  - `image_size` (optional; defaults to 3840×2160). Accepts `square`, `square_hd`, `portrait_4_3`, `portrait_16_9`, `landscape_4_3`, or `landscape_16_9`, or a custom `{"width": 3840, "height": 2160}` object. Custom dimensions must each be integers from 1 to 14,142.
  - `num_inference_steps` (integer 1–20; default 2)
  - `seed` (optional integer or `null`)
  - `guidance_scale` (number 0–20; default 5)
  - `num_images` (must be 1; default 1, one output per prepaid job)
  - `enable_safety_checker` (must be `true`; default `true`)
  - `output_format` (`jpeg` or `png`; default `jpeg`)
  - `style_name` (default `"(No style)"`; supported values: `"(No style)"`, `"Cinematic"`, `"Photographic"`, `"Anime"`, `"Manga"`, `"Digital Art"`, `"Pixel art"`, `"Fantasy art"`, `"Neonpunk"`, or `"3D Model"`)

Example request body:

```json
{
  "model": "sana-sprint",
  "input": {
    "prompt": "A tiny glass greenhouse on a mossy moon, soft sunrise, editorial photography",
    "image_size": {"width": 3840, "height": 2160},
    "output_format": "jpeg"
  }
}
```

## Job lifecycle

1. Check `GET /v2/models` and `GET /v2/billing/prices`; use this model only if it is enabled and the price is available.
2. Submit the request to `POST /v2/jobs` with a fresh `Idempotency-Key`. A successful submission returns `202 Accepted` and a job ID.
3. Poll `GET /v2/jobs/{job_id}` until `status` is `succeeded`, `failed`, or `submission_unknown`.
4. On success, inspect `artifacts` and download the generated image using its authenticated `download_url`. AI3D archives provider output in the job artifacts; this model does not create a 3D viewer link.

Fal lists pricing at $0.0025 per megapixel. AI3D estimates the reservation from the requested output dimensions, rounding up to a whole megapixel, and settles the final charge using Fal's reported billable units. The default 3840×2160 image reserves an estimate of 9 megapixels; check `/v2/billing/prices` for the current rate and estimate. The final charge may differ from the estimate.

When retrying a timed-out submission, resend the identical body with the same idempotency key. Never create a second job automatically when status is `submission_unknown`; provider acceptance and billing may need operator reconciliation.

## Provider reference

- <https://fal.ai/models/fal-ai/sana/sprint/llms.txt>