Models

At the moment, Skytells supports the following models on our general-purpose API:

Model	Vendor	Type	Description
TrueFusion	Skytells	Image Generation	Generate images from text.
TrueFusion Pro	Skytells	Image Generation	Generate images from text.
TrueFusion Max	Skytells	Image Generation	Generate images from text.
TrueFusion Pano	Skytells	Image Generation	Generate images from text.
TrueFusion Video	Skytells	Video Generation	Generate videos from text.
TrueFusion Video Pro	Skytells	Video Generation	Generate videos from text.
Imagen3	Google	Image Generation	Generate images from text.
Sana	NVIDIA	Image Generation	Generate images from text.

GET/v1/models

Available Models

This endpoint allows you to retrieve a paginated list of all models. By default, a maximum of ten models are shown per page.

This endpoint is public and useful for getting the list of models and their parameters, You may also use it to check the status of the model, if it is operational or not.

Optional attributes

Name
limit
Type
integer
Description
Limit the number of models returned.
Name
offset
Type
integer
Description
Offset the returned models.
Name
vendor
Type
object
Description
The vendor of the model.
Name
type
Type
string
Description
The type of the model.
Name
status
Type
string
Description
The status of the model. operational or offline.

Rate Limits

This endpoint is public and does not count towards your rate limits. However, you may still be rate limited by the public API limiter.

Request

GET

/v1/models

curl -G https://api.skytells.ai/v1/models \
  -H "x-api-key: {api_key}" \
  -d limit=10

Response

{
  "has_more": false,
  "models": [
    {
      "name": "TrueFusion",
      "description": "TrueFusion Standard",
      "namespace": "truefusion",
      "type": "image",
      "privacy": "public",
      "vendor": {
        "name": "Skytells",
        "description": "Skytells is a cutting-edge AI company at the forefront of technological innovation. With a mission to transform industries and empower businesses.",
        "image_url": "IMAGE_URL",
        "verified": true,
        "slug": "skytells",
        "metadata": null
      },
      "billable": true,
      "pricing": {
        "amount": 0.003,
        "currency": "USD",
        "unit": "image"
      },
      "status": "operational"
    }
  ]
}

TrueFusion

TrueFusion is a model that can generate images from text with a focus on realism and detail. To learn more about the TrueFusion family, please refer to the TrueFusion Model Card.

The TrueFusion model for image-generation is available in four variants:

TrueFusion: API name truefusion
TrueFusion Pro: API name truefusion-pro
TrueFusion Max: API name truefusion-max
TrueFusion Pano: API name truefusion-pano

While the TrueFusion models for videos are available in 2 variants:

TrueFusion Video: API name truefusion-video
TrueFusion Video Pro: API name truefusion-video-pro

Each variant has different parameters, inference time, and capabilities.

Model	Cost	Inference Time	Namespaces	Resolution
TrueFusion	$0.03 per image	~10 - 60 seconds	`truefusion`	1MP
TrueFusion Pro	$0.05 per image	~4 - 38 seconds	`truefusion-pro`	2MP
TrueFusion Max	$0.08 per image	~2 - 18 seconds	`truefusion-max`	4MP
TrueFusion Pano	$0.03 per image	~10 - 60 seconds	`truefusion-pano`	2MP
TrueFusion Video	$0.112 per second	~1 - 6 minutes	`truefusion-video`	720p
TrueFusion Video Pro	$0.196 per second	~1 - 5 minutes	`truefusion-video-pro`	1080p

Inference time is dependent on many factors, including the complexity of the prompt, the number of images, steps, and the size and aspect ratio of the image. However, Skytells runs all models on a custom-tuned H100 GPU, so the inference time is the same for all models.

You may reduce your billing costs by reducing the number of images, steps, and the size and aspect ratio of the image.

Parameters

Parameter	TrueFusion	TrueFusion Pro	TrueFusion Max	TrueFusion Pano
`prompt`	✓	✓	✓	✓
`width`	✓	✓	✓	✓
`height`	✓	✓	✓	✓
`num_inference_steps`	✓	✓	✓	✓
`guidance_scale`	✓	✓	✓	✓
`seed`	✓	✓	✓	✓
`num_images`		✓	✓	✓
`output_format`		✓	✓	✓
`output_quality`		✓	✓	✓