Yapper Docs
Models & Inputs

Image generation

Image models with per-resolution pricing and input fields.

POST /api/v1/processes with "type": "image-generation". Pick any model — the API does not plan-gate models. GET /api/v1/models returns the same capabilities at runtime.

Input fields

FieldTypeRequiredDescription
promptstringyesWhat to generate.
aspectRatiostringnoOne of the model's aspect ratios, e.g. "16:9". Use "adaptive" when advertised to follow the primary image/video reference.
imageResolutionstringnoOne of the model's resolutions, e.g. "1080" or "4k".
referenceImageUrlsstring[]noPlain https image URLs to guide/edit with. External URLs are imported into your asset library first (validated + stored); team asset URLs pass through untouched.
referenceImages({assetId} | {url})[]noImage references. Send either {assetId} for a team asset or {url} for a public URL to import. Legacy {docId, url} objects remain accepted.
promptUpsamplingbooleannoLet the provider expand the prompt before generating.
disableAutoRetriesbooleannoDisable automatic same-model retries. Defaults to false, so automatic retries are enabled.
maskUrlstringnoInpainting mask URL (white = regenerate), for edits with GPT Image 2, 2.5 and 2.6 when referenceImages are given. Other models ignore it.
markupUrlstringnoAnnotation layer URL for GPT Image 2.6 edits: a transparent PNG the size of the first reference image, marking what to change. Describe the change in the prompt. Not combined with maskUrl.
customSettingsobjectnoThe model's own settings, the same ones the app's settings menu shows (for example GPT Image 2.6 quality, complexity and background: transparent). Only the settings listed in the model's capabilities.customSettings are accepted. Omitted settings keep their default.
outputNamestringnoName for the output asset, up to 80 characters. It is stored as a lowercase hyphenated slug ("ORD-17 Concept A" becomes ord-17-concept-a), so use metadata for exact identifiers.

Referencing media: referenceImageUrls takes plain https image URLs. The normalized referenceImages, referenceVideos, and referenceAudios fields accept either {assetId} for an existing team asset or {url} for public media, which is imported automatically. Provider-native asset:// URIs are internal and are not accepted here. Legacy {docId, url} objects remain accepted for backwards compatibility.

Example

Add "dryRun": true to get the exact credit cost without starting, and always send an Idempotency-Key header on real starts.

POST /api/v1/processes
{
  "type": "image-generation",
  "model": "soul-2",
  "input": {
    "prompt": "A polished product photo of a stainless steel water bottle",
    "aspectRatio": "3:4",
    "imageResolution": "1080"
  }
}

Models at a glance

Credits are per generated output; ranges span the supported resolutions/lengths (exact prices in each model's section).

ModelCompanyCreditsResolutionsMax ref images
gpt-image-2.6-flareopenai9–29720, 1080, 2k, 4k16
gpt-image-2.6-sunburstopenai9–29720, 1080, 2k, 4k16
gpt-image-2.5-flareopenai9–29720, 1080, 2k, 4k16
gpt-image-2.5-sunburstopenai9–29720, 1080, 2k, 4k16
nano-banana-progoogle21–36720, 1080, 2k, 4k14
nano-banana-2.1google7–211080, 2k, 4k14
nano-banana-2google11–23720, 1080, 2k, 4k14
nano-banana-2-litegoogle9108014
gemini-imagegoogle8–10720, 10803
gpt-image-2openai10–20720, 1080, 2k, 4k16
seedream-v4.5bytedance8–201080, 2k, 4k10
seedream-v5.0-probytedance5–101080, 2k10
seedream-v5.0-litebytedance202k, 4k14
seedream-v5.5bytedance9–181080, 2k10
grok-imagine-image-2.0xai13–181080, 2k3
grok-imagine-imagexai61080, 2k3
flux-2-problack-forest-labs10–201080, 2k8
flux-3-imageblack-forest-labs10–1301080, 2k, 4k10
kling-image-4.0kuaishou14–342k, 4k, 8k10
minimax-h3-imageminimax6–161080, 2k, 4k16
soul-2higgsfield22k0
gpt-image-2-highopenai30–80720, 1080, 2k, 4k16

GPT Image 2.6 Flare (gpt-image-2.6-flare)

Fast, high-quality image generation and editing from OpenAI.

Credits720: 9 · 1080: 12 · 2k: 17 · 4k: 29
Aspect ratiosauto, 1:1, 16:9, 21:9, 9:16, 3:4, 4:5, 4:3, 5:4, 3:2, 2:3 (default 4:3)
Resolutions720, 1080, 2k, 4k
customSettingsquality: low, medium, high, xhigh, max (default high)
complexity: low, medium, high (default high)
background: auto, opaque, transparent (default auto)
Reference imagesup to 16

GPT Image 2.6 Sunburst (gpt-image-2.6-sunburst)

OpenAI's most precise image model for detailed visuals and consistent edits.

Credits720: 9 · 1080: 12 · 2k: 17 · 4k: 29
Aspect ratiosauto, 1:1, 16:9, 21:9, 9:16, 3:4, 4:5, 4:3, 5:4, 3:2, 2:3 (default 4:3)
Resolutions720, 1080, 2k, 4k
customSettingsquality: low, medium, high, xhigh, max (default high)
complexity: low, medium, high (default high)
background: auto, opaque, transparent (default auto)
Reference imagesup to 16

GPT Image 2.5 Flare (gpt-image-2.5-flare)

Fast, high-quality image generation and editing from OpenAI.

Credits720: 9 · 1080: 12 · 2k: 17 · 4k: 29
Aspect ratiosauto, 1:1, 16:9, 21:9, 9:16, 3:4, 4:5, 4:3, 5:4, 3:2, 2:3 (default 4:3)
Resolutions720, 1080, 2k, 4k
customSettingsquality: low, medium, high, xhigh, max (default high)
background: auto, opaque, transparent (default auto)
Reference imagesup to 16

GPT Image 2.5 Sunburst (gpt-image-2.5-sunburst)

OpenAI's most precise image model for detailed visuals and consistent edits.

Credits720: 9 · 1080: 12 · 2k: 17 · 4k: 29
Aspect ratiosauto, 1:1, 16:9, 21:9, 9:16, 3:4, 4:5, 4:3, 5:4, 3:2, 2:3 (default 4:3)
Resolutions720, 1080, 2k, 4k
customSettingsquality: low, medium, high, xhigh, max (default high)
background: auto, opaque, transparent (default auto)
Reference imagesup to 16

Nano Banana Pro (nano-banana-pro)

Google's advanced image generation and editing model.

Credits720: 21 · 1080: 21 · 2k: 21 · 4k: 36
Aspect ratiosauto, 1:1, 16:9, 21:9, 9:16, 3:4, 4:5, 4:3, 5:4, 3:2, 2:3 (default 1:1)
Resolutions720, 1080, 2k, 4k
Reference imagesup to 14

Nano Banana 2.1 (nano-banana-2.1)

Google's newest image model, with sharper text and more consistent edits.

Credits1080: 7 · 2k: 10 · 4k: 21
Aspect ratiosauto, 1:1, 16:9, 21:9, 9:16, 3:4, 4:5, 4:3, 5:4, 3:2, 2:3 (default 1:1)
Resolutions1080, 2k, 4k
Reference imagesup to 14

Nano Banana 2 (nano-banana-2)

Google's previous all-purpose image model.

Credits720: 11 · 1080: 12 · 2k: 18 · 4k: 23
Aspect ratiosauto, 1:1, 16:9, 21:9, 9:16, 3:4, 4:5, 4:3, 5:4, 3:2, 2:3 (default 1:1)
Resolutions720, 1080, 2k, 4k
Reference imagesup to 14

Nano Banana 2 Lite (nano-banana-2-lite)

Google's fastest image model.

Credits1080: 9
Aspect ratiosauto, 1:1, 16:9, 21:9, 9:16, 3:4, 4:5, 4:3, 5:4, 3:2, 2:3 (default 1:1)
Resolutions1080
Reference imagesup to 14

Nano Banana (gemini-image)

Google's legacy image model with quick generation times.

Credits720: 8 · 1080: 10
Aspect ratiosauto, 1:1, 16:9, 21:9, 9:16, 3:4, 4:5, 4:3, 5:4, 3:2, 2:3 (default 16:9)
Resolutions720, 1080
Reference imagesup to 3

GPT Image 2 (gpt-image-2)

Strong image generation and editing model that's especially good at text and typography, with precise instruction following and balanced cost.

Credits720: 10 · 1080: 10 · 2k: 10 · 4k: 20
Aspect ratiosauto, 1:1, 16:9, 21:9, 9:16, 3:4, 4:5, 4:3, 5:4, 3:2, 2:3 (default 4:3)
Resolutions720, 1080, 2k, 4k
customSettingsquality: low, medium, high (default medium)
Reference imagesup to 16

Seedream-V4.5 (seedream-v4.5)

Bytedance's advanced image generation model.

Credits1080: 8 · 2k: 10 · 4k: 20
Aspect ratiosauto, 1:1, 16:9, 9:16, 3:4, 4:5, 4:3, 5:4 (default 16:9)
Resolutions1080, 2k, 4k
Reference imagesup to 10

Seedream-V5.0 Pro (seedream-v5.0-pro)

Bytedance's flagship Seedream image generation and editing model.

Credits1080: 5 · 2k: 10
Aspect ratiosauto, 1:1, 16:9, 9:16, 4:3, 5:4, 3:4, 4:5, 3:2, 2:3 (default 1:1)
Resolutions1080, 2k
Reference imagesup to 10

Seedream-V5.0 Lite (seedream-v5.0-lite)

Bytedance's fast next-generation Seedream image generation and editing model.

Credits2k: 20 · 4k: 20
Aspect ratiosauto, 1:1, 16:9, 9:16, 4:3, 5:4, 3:4, 4:5, 3:2, 2:3 (default 1:1)
Resolutions2k, 4k
Reference imagesup to 14

Seedream-V5.5 (seedream-v5.5)

ByteDance's newest Seedream model, with more consistent edits, more realistic images and more accurate text.

Credits1080: 9 · 2k: 18
Aspect ratiosauto, 1:1, 16:9, 9:16, 4:3, 5:4, 3:4, 4:5, 3:2, 2:3 (default 1:1)
Resolutions1080, 2k
customSettingsbackground: opaque, transparent (default opaque)
Reference imagesup to 10

Grok Image 2 (grok-imagine-image-2.0)

xAI's latest image generation and editing model, with stronger typography, layout, and instruction following.

Credits1080: 13 · 2k: 18
Aspect ratiosauto, 1:1, 16:9, 9:16, 4:3, 3:4, 3:2, 2:3, 2:1, 1:2 (default 1:1)
Resolutions1080, 2k
customSettingsquality: low, medium (default medium)
Reference imagesup to 3
Max prompt length8000

Grok Image (grok-imagine-image)

xAI's image generation and editing model.

Credits1080: 6 · 2k: 6
Aspect ratiosauto, 1:1, 16:9, 9:16, 4:3, 3:4, 3:2, 2:3 (default 1:1)
Resolutions1080, 2k
Reference imagesup to 3

Flux.2 Pro (flux-2-pro)

Flux's advanced image generation model.

Credits1080: 10 · 2k: 20
Aspect ratiosauto, 1:1, 16:9, 9:16, 3:4, 4:5, 4:3, 5:4, 2:3, 3:2 (default 16:9)
Resolutions1080, 2k
Reference imagesup to 8

FLUX.3 Image (flux-3-image)

Black Forest Labs' flagship image model, with native 2K and 4K output, precise layout control, and sharp text.

Credits1080: 10 · 2k: 21 · 4k: 130
Aspect ratiosauto, 1:1, 16:9, 21:9, 9:16, 3:4, 4:5, 4:3, 5:4, 3:2, 2:3, 2:1, 1:2 (default 16:9)
Resolutions1080, 2k, 4k
Reference imagesup to 10

Kling Image 4.0 (kling-image-4.0)

Kuaishou's image model, with 2K, 4K and 8K output and up to 10 reference images.

Credits2k: 14 · 4k: 20 · 8k: 34
Aspect ratiosauto, 1:1, 16:9, 9:16, 4:3, 3:4, 3:2, 2:3, 21:9 (default 16:9)
Resolutions2k, 4k, 8k
Reference imagesup to 10

MiniMax H3 Image (minimax-h3-image)

MiniMax's open-weight image generation and editing model, with up to 16 references and fast 2K edits.

Credits1080: 6 · 2k: 6 · 4k: 16
Aspect ratiosauto, 1:1, 16:9, 9:16, 4:3, 3:4, 3:2, 2:3, 21:9 (default 16:9)
Resolutions1080, 2k, 4k
Reference imagesup to 16

Soul 2 (soul-2)

Higgsfield's photographic text-to-image model with an editorial, fashion-shoot look.

Credits2k: 2
Aspect ratios1:1, 16:9, 9:16, 4:3, 3:4, 3:2, 2:3 (default 3:4)
Resolutions2k
Reference imagesup to 0

GPT Image 2 High Effort (gpt-image-2-high)

High-effort GPT Image 2 variant for stronger fidelity and more deliberate final outputs.

Credits720: 30 · 1080: 30 · 2k: 50 · 4k: 80
Aspect ratiosauto, 1:1, 16:9, 21:9, 9:16, 3:4, 4:5, 4:3, 5:4, 3:2, 2:3 (default 4:3)
Resolutions720, 1080, 2k, 4k
Reference imagesup to 16

On this page