Editing Models - Parameters Reference
Google
Gemini 3.1 Flash (Nano Banana 2)
Model ID: model_google-gemini-3-1-flash
| Input | Label | Type | Default | Min | Max | Allowed Values | Notes |
|---|---|---|---|---|---|---|---|
prompt | Prompt | string | None | None | 4096 | None | Required. Text prompt for image generation or editing. |
referenceImages | Reference images | file_array | None | None | 14 | None | Up to 14 reference images for style or content guidance. |
aspectRatio | Aspect Ratio | string | auto | None | None | 21:9, 16:9, 3:2, 4:3, 5:4, 1:1, 4:5, 3:4, 2:3, 9:16, auto | Aspect ratio for the generated image. |
resolution | Resolution | string | 1K | None | None | 512, 1K, 2K, 4K | Target resolution for the output. |
useGoogleSearch | Use Google Search | boolean | false | None | None | None | Use Google Search to find more information about the prompt. |
numOutputs | Image Count | number | 1 | 1 | 4 | None | Number of images to generate (1-4). |
seed | Seed | number | None | 0 | 2147483647 | None | Use for reproducible results. Leave blank for random. |
Gemini 3.0 Pro
Model ID: model_google-gemini-pro-image-editing
| Input | Label | Type | Default | Min | Max | Allowed Values | Notes |
|---|---|---|---|---|---|---|---|
referenceImages | Reference images | assetId[] | – | – | 14 | – | Required. Up to 14 reference images |
prompt | Instructions | string | – | – | – | – | Required text prompt (max 4096 chars) |
aspectRatio | Aspect Ratio | string | auto | – | – | 21:9, 16:9, 3:2, 4:3, 5:4, 1:1, 4:5, 3:4, 2:3, 9:16, auto | Defines output aspect ratio |
resolution | Resolution | string | 2K | – | – | 1K, 2K, 4K | Output resolution |
numOutputs | Image Count | number | 1 | 1 | 4 | – | Number of images to generate |
seed | Seed | number | – | 0 | 2147483647 | – | Optional seed for reproducibility |
Gemini 2.5 🍌
Model ID: model_google-gemini-2-5-flash-image-editing
| Input | Label | Type | Default | Min | Max | Allowed Values | Notes |
|---|---|---|---|---|---|---|---|
referenceImages | Reference Images | assetId[] | – | – | 6 | – | Required reference images for style or content guidance |
prompt | Instructions | string | – | – | – | – | Required text prompt for image generation/editing (max 4096 chars) |
aspectRatio | Aspect Ratio | string | auto | – | – | 21:9, 16:9, 3:2, 4:3, 5:4, 1:1, 4:5, 3:4, 2:3, 9:16, auto | Output aspect ratio |
numOutputs | Image Count | number | 1 | 1 | 4 | – | Number of images to generate |
seed | Seed | number | – | 0 | 2147483647 | – | Optional reproducibility seed |
Flux Kontext
Model ID: model_flux-kontext-editing
| Input | Label | Type | Default | Min | Max | Allowed Values | Notes |
|---|---|---|---|---|---|---|---|
loras | Flux Kontext LoRA | modelId[] | [] | – | – | – | List of LoRA model IDs or URLs (max 10) |
lorasScale | LoRA Scale | number[] | [1.0] | 0.0 | 2.0 | – | Scales for LoRA weights (max 10) |
referenceImages | Reference Images | assetId[] | – | – | – | – | Required reference images for style/content guidance (max 10) |
prompt | Instructions | string | – | – | – | – | Required text prompt for generation or editing (max 4096 chars) |
numOutputs | Image Count | number | 1 | 1 | 4 | – | Number of images to generate |
aspectRatio | Aspect Ratio | string | auto | – | – | 21:9, 16:9, 3:2, 4:3, 1:1, 2:3, 3:4, 9:16, 9:21, auto | Output aspect ratio |
quality | Quality | string | high | – | – | high, medium | Image generation quality |
guidanceScale | Guidance | number | 2.5 | 1.0 | 20.0 | – | How closely to follow the prompt |
seed | Seed | number | – | 0 | 2147483647 | – | Optional random seed for reproducibility |
Bytedance
Seedream 5.0 Lite
Model ID: model_bytedance-seedream-5-0-lite
| Input | Label | Type | Default | Min | Max | Allowed Values | Notes |
|---|---|---|---|---|---|---|---|
referenceImages | Reference Images | assetId[] | [] | None | 14 | None | Required. One to ten images. First image is the base image to edit |
prompt | Instructions | string | None | None | 2048 | None | Required text instructions for the edit |
aspectRatio | Aspect Ratio | string | match_input_image | None | None | match_input_image, 21:9, 16:9, 3:2, 4:3, 1:1, 3:4, 2:3, 9:16 | Output image aspect ratio |
width | Width | number | 2048 | 1024 | 4704 | None | The width of the generated images, must be a multiple of 16. |
height | Height | number | 2048 | 1024 | 4704 | None | The height of the generated images, must be a multiple of 16. |
sequentialImageGeneration | Generate Multiple Outputs | string | disabled | – | – | disabled, auto | Controls group image generation mode |
maxImages | Image Count | number | 1 | 1 | 15 | – | Maximum images to generate when sequential mode is on |
Seedream 4.5
Model ID: model_bytedance-seedream-4-5-editing
| Input | Label | Type | Default | Min | Max | Allowed Values | Notes |
|---|---|---|---|---|---|---|---|
referenceImages | Reference Images | assetId[] | [] | None | 14 | None | Required. One to ten images. First image is the base image to edit |
prompt | Instructions | string | None | None | 2048 | None | Required text instructions for the edit |
aspectRatio | Aspect Ratio | string | match_input_image | None | None | match_input_image, 21:9, 16:9, 3:2, 4:3, 1:1, 3:4, 2:3, 9:16 | Output image aspect ratio |
sequentialImageGeneration | Generate Multiple Outputs | string | disabled | – | – | disabled, auto | Controls group image generation mode |
maxImages | Image Count | number | 1 | 1 | 15 | – | Maximum images to generate when sequential mode is on |
Seedream 4
Model ID: model_bytedance-seedream-4-editing
| Input | Label | Type | Default | Min | Max | Allowed Values | Notes |
|---|---|---|---|---|---|---|---|
referenceImages | Reference Images | assetId[] | [] | – | 10 | – | Required. 1–10 input images for single or multi-reference generation |
prompt | Instructions | string | – | – | – | – | Required text prompt for image generation (max 2048 chars) |
sequentialImageGeneration | Generate Multiple Outputs | string | disabled | – | – | disabled, auto | Controls group image generation mode |
maxImages | Image Count | number | 1 | 1 | 15 | – | Maximum images to generate when sequential mode is on |
aspectRatio | Aspect Ratio | string | match_input_image | – | – | match_input_image, 21:9, 16:9, 3:2, 4:3, 1:1, 3:4, 2:3, 9:16 | Output aspect ratio |
size | Quality | string | 4K | – | – | 4K, 2K, 1K | Output resolution quality |
OpenAI
GPT Image 1.5
Model ID: model_openai-gpt-image-1-5-editing
| Input | Label | Type | Default | Min | Max | Allowed Values | Notes |
|---|---|---|---|---|---|---|---|
referenceImages | Reference Images | assetId[] | None | None | 10 | None | Required. First image is the base image to edit |
prompt | Instructions | string | None | None | 32000 | None | Required text instructions for editing |
numOutputs | Image Count | number | 1 | 1 | 10 | None | Number of images generated |
aspectRatio | Aspect Ratio | string | auto | None | None | auto, 1:1, 3:2, 2:3 | Output image aspect ratio |
inputFidelity | Input Fidelity | string | high | None | None | low, high | Preserve input image details |
quality | Quality | string | high | None | None | high, medium, low | Output image quality |
background | Background | string | auto | None | None | transparent, opaque, auto | Background handling |
GPT Image 1
Model ID: model_openai-gpt-image-1-editing
| Input | Label | Type | Default | Min | Max | Allowed Values | Notes |
|---|---|---|---|---|---|---|---|
referenceImages | Reference Images | assetId[] | – | – | 4 | – | Required reference images for style or content guidance |
prompt | Instructions | string | – | – | – | – | Required text prompt for image generation/editing (max 32000 chars) |
numOutputs | Number of Images | number | 1 | 1 | 4 | – | Number of images to generate |
aspectRatio | Aspect Ratio | string | auto | – | – | 3:2, 1:1, 2:3, auto | Aspect ratio for output images |
inputFidelity | Input Fidelity | string | high | – | – | low, high | Controls how closely to preserve details from input images |
quality | Quality | string | high | – | – | high, medium, low | Generation quality level |
Qwen
Qwen Edit 2511
Model ID: qwen-image-edit-2511
| Input | Label | Type | Default | Min | Max | Allowed Values | Notes |
|---|---|---|---|---|---|---|---|
modelId | LoRA Model | model | "" | None | None | None | A single LoRA model ID. If provided, it will be added to the loras array. |
loras | Qwen LoRA | model_array | [] | None | None | None | List of one or more LoRA model IDs (max 6). |
lorasScale | LoRA Scale | number_array | None | 0.0 | 2.0 | None | Scales for the LoRA weights (step 0.05). |
prompt | Instructions | string | None | None | 2048 | None | Required. Text prompt for image editing. |
referenceImages | Reference Images | file_array | None | None | None | None | Required. Reference images for editing. |
numOutputs | Image Count | number | 1 | 1 | 4 | None | Number of images to generate. |
numInferenceSteps | Inference Steps | number | 28 | 1 | 100 | None | The number of denoising steps. |
width | Width | number | None | 128 | 2048 | None | The width of the generated images, must be a multiple of 16. |
height | Height | number | None | 128 | 2048 | None | The height of the generated images, must be a multiple of 16. |
guidance | Guidance | number | 4.0 | 1.0 | 10.0 | None | Controls how closely the generated image follows the prompt (step 0.1). |
seed | Seed | number | None | 0 | 2147483647 | None | Used to reproduce previous results. |
Qwen Edit 2509
Model ID: qwen-image-edit-2509
| Input | Label | Type | Default | Min | Max | Allowed Values | Notes |
|---|---|---|---|---|---|---|---|
modelId | LoRA Model | model | "" | None | None | None | A single LoRA model ID. If provided, it will be added to the loras array. |
loras | Qwen LoRA | model_array | [] | None | None | None | List of one or more LoRA model IDs (max 6). |
lorasScale | LoRA Scale | number_array | None | 0.0 | 2.0 | None | Scales for the LoRA weights (step 0.05). |
prompt | Instructions | string | None | None | 2048 | None | Required. Text prompt for image editing. |
referenceImages | Reference Images | file_array | None | None | None | None | Required. Reference images for editing. |
numOutputs | Image Count | number | 1 | 1 | 4 | None | Number of images to generate. |
numInferenceSteps | Inference Steps | number | 28 | 1 | 100 | None | The number of denoising steps. |
width | Width | number | None | 128 | 2048 | None | The width of the generated images, must be a multiple of 16. |
height | Height | number | None | 128 | 2048 | None | The height of the generated images, must be a multiple of 16. |
guidance | Guidance | number | 4.0 | 1.0 | 10.0 | None | Controls how closely the generated image follows the prompt (step 0.1). |
seed | Seed | number | None | 0 | 2147483647 | None | Used to reproduce previous results. |
Qwen Edit Multiangle 2511
Model ID: model_qwen-edit-multiangle-2511
| Input | Label | Type | Default | Min | Max | Allowed Values | Notes |
|---|---|---|---|---|---|---|---|
images | Image | assetId[] | None | 1 | 1 | None | Required. Input image to edit |
horizontalAngle | Horizontal Angle (Azimuth) | number | 0 | 0 | 360 | None | 0 front, 90 right, 180 back, 270 left |
verticalAngle | Vertical Angle (Elevation) | number | 0 | -30 | 90 | None | -30 low angle, 0 eye level, 90 top down |
zoom | Zoom | number | 5 | 0 | 10 | None | 0 wide shot, 10 close up |
loraScale | LoRA Scale | number | 1.5 | 0 | 4 | None | Controls camera effect strength |
numOutputs | Image Count | number | 1 | 1 | 4 | None | Number of images generated |
numInferenceSteps | Steps | number | 28 | 1 | 50 | None | Higher values increase quality |
guidanceScale | Guidance Scale | number | 4.5 | 1 | 20 | None | Higher values follow the prompt more closely |
acceleration | Acceleration | string | none | None | None | none, regular | Enables faster generation |
seed | Seed | number | None | 0 | 2147483647 | None | Optional reproducibility seed |
Qwen Edit Plus
Model ID: model_qwen-image-edit
| Input | Label | Type | Default | Min | Max | Allowed Values | Notes |
|---|---|---|---|---|---|---|---|
modelId | LoRA Model | model | "" | None | None | None | A single LoRA model ID. If provided, it will be added to the loras array. |
loras | Qwen LoRA | model_array | [] | None | None | None | List of one or more LoRA model IDs (max 6). |
lorasScale | LoRA Scale | number_array | None | 0.0 | 2.0 | None | Scales for the LoRA weights (step 0.05). |
prompt | Instructions | string | None | None | 2048 | None | Required. Text prompt for image editing. |
referenceImages | Reference Image | file | None | None | None | None | Required. Input image for editing. |
numOutputs | Image Count | number | 1 | 1 | 4 | None | Number of images to generate. |
numInferenceSteps | Inference Steps | number | 28 | 1 | 100 | None | The number of denoising steps. |
width | Width | number | None | 128 | 2048 | None | Width of generated images (multiple of 16). |
height | Height | number | None | 128 | 2048 | None | Height of generated images (multiple of 16). |
guidance | Guidance | number | 4.0 | 1.0 | 10.0 | None | Controls how closely the generated image follows the prompt (step 0.1). |
seed | Seed | number | None | 0 | 2147483647 | None | Used to reproduce previous results. |
Runway Gen 4
Model ID: model_runway-gen4-image-editing
| Input | Label | Type | Default | Min | Max | Allowed Values | Notes |
|---|---|---|---|---|---|---|---|
referenceImages | Reference Images | assetId[] | [] | – | 3 | – | Required. Up to 3 input reference images |
prompt | Instructions | string | – | – | – | – | Required text prompt (max 2048 chars) |
resolution | Quality | string | 1080p | – | – | 720p, 1080p | Output image resolution |
aspectRatio | Aspect Ratio | string | 16:9 | – | – | 21:9, 16:9, 4:3, 3:2, 1:1, 3:4, 9:16 | Output aspect ratio |
seed | Seed | number | – | 0 | 2147483647 | – | Optional reproducibility seed |
P-Image Edit
Model ID: model_p-image-editing
| Input | Label | Type | Default | Min | Max | Allowed Values | Notes |
|---|---|---|---|---|---|---|---|
prompt | Instructions | string | None | None | 2048 | None | Required. Describe the edit task clearly. You can reference images as image 1, image 2, etc |
referenceImages | Reference Images | assetId[] | None | None | 10 | None | Required. First image is the main image to edit |
turbo | Turbo | boolean | true | None | None | None | Enables faster execution. Disable for complex edits |
aspectRatio | Aspect Ratio | string | match_input_image | None | None | 16:9, 3:2, 4:3, 1:1, 3:4, 2:3, 9:16, match_input_image | Output image aspect ratio |
seed | Seed | number | None | None | None | None | Optional seed for reproducible results |
LongCat Image Editing
Model ID: model_longcat-image-editing
| Input | Label | Type | Default | Min | Max | Allowed Values | Notes |
|---|---|---|---|---|---|---|---|
referenceImage | Reference Image | assetId | None | None | None | None | Required. Base image to edit |
prompt | Instructions | string | None | None | 2048 | None | Required text instructions describing the edit |
numInferenceSteps | Steps | number | 28 | 1 | 50 | None | Number of inference steps |
guidanceScale | Guidance | number | 4.5 | 1 | 20 | None | Higher values follow the prompt more closely |
numOutputs | Image Count | number | 1 | 1 | 4 | None | Number of images generated |
seed | Seed | number | None | 0 | 2147483647 | None | Optional reproducibility seed |
Black Forest Labs
Flux 2 Max Edit
Model ID: model_bfl-flux-2-max-editing
| Input | Label | Type | Default | Min | Max | Allowed Values | Notes |
|---|---|---|---|---|---|---|---|
referenceImages | Reference Images | assetId[] | None | None | 8 | None | Required. Reference images for style or content guidance |
prompt | Instructions | string | None | None | 2048 | None | Required text instructions for image editing |
aspectRatio | Aspect Ratio | string | 4:3 | None | None | match_input_image, 16:9, 3:2, 4:3, 5:4, 1:1, 4:5, 3:4, 2:3, 9:16 | Output image aspect ratio |
resolution | Resolution | string | 1 MP | None | None | match_input_image, 0.5 MP, 1 MP, 2 MP, 4 MP | Output resolution in megapixels |
seed | Seed | number | None | None | None | None | Optional reproducibility seed |
Flux 2 Flex Edit
Model ID: model_bfl-flux-2-flex-editing
| Input | Label | Type | Default | Min | Max | Allowed Values | Notes |
|---|---|---|---|---|---|---|---|
referenceImages | Reference Images | assetId[] | None | None | 10 | None | Required. Reference images for guidance |
prompt | Instructions | string | None | None | 2048 | None | Required text instructions |
aspectRatio | Aspect Ratio | string | 1:1 | None | None | match_input_image, 16:9, 3:2, 4:3, 5:4, 1:1, 4:5, 3:4, 2:3, 9:16 | Output image aspect ratio |
resolution | Resolution | string | 1 MP | None | None | match_input_image, 0.5 MP, 1 MP, 2 MP, 4 MP | Output resolution |
promptUpsampling | Prompt Upsampling | boolean | false | None | None | None | Automatically rewrites the prompt |
steps | Steps | number | 30 | 1 | 50 | None | Number of inference steps |
guidance | Guidance | number | 4.5 | 1.5 | 10 | None | Higher values follow the prompt more closely |
seed | Seed | number | None | None | None | None | Optional seed |
Flux 2 Pro Edit
Model ID: model_bfl-flux-2-pro-editing
| Input | Label | Type | Default | Min | Max | Allowed Values | Notes |
|---|---|---|---|---|---|---|---|
referenceImages | Reference Images | assetId[] | None | None | 8 | None | Required. Reference images for style or content |
prompt | Instructions | string | None | None | 2048 | None | Required text instructions |
aspectRatio | Aspect Ratio | string | 1:1 | None | None | match_input_image, 16:9, 3:2, 4:3, 5:4, 1:1, 4:5, 3:4, 2:3, 9:16 | Output image aspect ratio |
resolution | Resolution | string | 1 MP | None | None | match_input_image, 0.5 MP, 1 MP, 2 MP, 4 MP | Output resolution |
promptUpsampling | Prompt Upsampling | boolean | false | None | None | None | Automatically enhances the prompt |
seed | Seed | number | None | None | None | None | Optional reproducibility seed |
Flux 2 Turbo Edit
Model ID: model_bfl-flux-2-turbo-editing
| Input | Label | Type | Default | Min | Max | Allowed Values | Notes |
|---|---|---|---|---|---|---|---|
referenceImages | Reference Images | assetId[] | None | None | 4 | None | Required. Reference images for style or content guidance |
prompt | Instructions | string | None | None | 2048 | None | Required text instructions for image editing |
imageSize | Image Size | string | landscape_4_3 | None | None | portrait_16_9, portrait_4_3, square_hd, landscape_4_3, landscape_16_9 | Output image size preset |
guidance | Guidance | number | 2.5 | 1.5 | 10 | None | Higher values follow the prompt more closely |
numOutputs | Image Count | number | 1 | 1 | 4 | None | Number of images generated |
promptUpsampling | Prompt Upsampling | boolean | false | None | None | None | Automatically rewrites the prompt for more creative results |
seed | Seed | number | None | None | None | None | Optional reproducibility seed |
Sourceful
Riverflow V2 Fast
Model ID: model_sourceful-riverflow-2-0-fast
| Input | Label | Type | Default | Min | Max | Allowed Values | Notes |
|---|---|---|---|---|---|---|---|
prompt | Instruction | string | None | None | 2048 | None | Required generation instruction |
referenceImages | Reference Images | assetId[] | None | None | 4 | None | Optional images to edit |
superResolutionRefs | Super Resolution Refs | assetId[] | None | None | 4 | None | Optional references for detail enhancement |
resolution | Resolution | string | 1K | None | None | 1K, 2K | Target resolution |
aspectRatio | Aspect Ratio | string | auto | None | None | auto, 21:9, 16:9, 3:2, 4:3, 5:4, 1:1, 4:5, 3:4, 2:3, 9:16 | Output aspect ratio |
fontUrls | Font Urls | string[] | None | None | 2 | None | Optional font URLs. Up to 2 |
fontTexts | Font Texts | string[] | None | None | None | None | Text strings matching fontUrls |
transparency | Transparency | boolean | false | None | None | None | Enables transparent background when supported |
enhancePrompt | Enhance Prompt | boolean | false | None | None | None | Allows model to enhance the instruction |
maxIterations | Max Iterations | number | 3 | 1 | 3 | None | Number of internal reasoning iterations |
Riverflow V2 REFSR
Model ID: model_sourceful-riverflow-2-0-refsr
| Input | Label | Type | Default | Min | Max | Allowed Values | Notes |
|---|---|---|---|---|---|---|---|
referenceImage | Reference Image | assetId | None | None | None | None | Required. Base image to apply super resolution to |
superResolutionReferences | Super Resolution References | assetId[] | None | None | 4 | None | Required. Reference images for super resolution guidance |
Riverflow V2 Pro
Model ID: model_sourceful-riverflow-2-0-pro
| Input | Label | Type | Default | Min | Max | Allowed Values | Notes |
|---|---|---|---|---|---|---|---|
prompt | Instruction | string | None | None | 2048 | None | Required generation instruction |
referenceImages | Reference Images | assetId[] | None | None | 10 | None | Optional images to edit |
superResolutionRefs | Super Resolution Refs | assetId[] | None | None | 4 | None | Optional references for detail enhancement |
resolution | Resolution | string | 1K | None | None | 1K, 2K, 4K | Target resolution |
aspectRatio | Aspect Ratio | string | auto | None | None | auto, 21:9, 16:9, 3:2, 4:3, 5:4, 1:1, 4:5, 3:4, 2:3, 9:16 | Output aspect ratio |
fontUrls | Font Urls | string[] | None | None | 2 | None | Optional font URLs. Up to 2 |
fontTexts | Font Texts | string[] | None | None | None | None | Text strings matching fontUrls |
transparency | Transparency | boolean | false | None | None | None | Enables transparent background when supported |
enhancePrompt | Enhance Prompt | boolean | false | None | None | None | Allows model to enhance the instruction |
maxIterations | Max Iterations | number | 3 | 1 | 3 | None | Number of internal reasoning iterations |
Physic Edit
Model ID: model_physic-edit
| Input | Label | Type | Default | Min | Max | Allowed Values | Notes |
|---|---|---|---|---|---|---|---|
image | Reference Image | file | None | None | None | None | Required. Input image to edit with physics-aware transformations. |
prompt | Instructions | string | None | None | 2048 | None | Required. Editing instructions describing the physical transformation to apply. |
negativePrompt | Negative Prompt | string | "" | None | 2048 | None | Description of unwanted elements in the output. |
numInferenceSteps | Inference Steps | number | 40 | 1 | 50 | None | Number of denoising steps. |
cfgScale | Guidance | number | 4 | 1 | 20 | None | Classifier-free guidance scale (step 0.5). |
seed | Seed | number | None | 0 | 2147483647 | None | Random seed for reproducibility. |
Updated 3 days ago