ComfyUI-Kie-API

KIE_Flux2_I2I

Image-to-image generation node with a Pro or Flex model choice, accepting 1–8 input images through a ComfyUI batch.

Kie/Api · 1 input · 5 parameters · 1 output

KIE_Gemini3Pro_LLM

Experimental text-generation node using Gemini chat completions, with role selection, image/video/audio inputs, optional reasoning output, and a Google Search toggle.

Kie/Api · 3 inputs · 10 parameters · 3 outputs

KIE_GetRemainingCredits

Check remaining KIE account credits using the API key from config/kiekey.txt. - log (BOOLEAN, optional): Enable console logging (default: true). - data (STRING): Raw JSON response from the credits endpoint. - creditsremaining (INT): Parsed remaining credit balance. - This node is useful as a first smoke test after setup. - If the API key is missing or invalid, the node raises an error.

Kie/Api · 0 inputs · 1 parameter · 2 outputs

KIE_GPTImage2_ImageToImage

GPT Image image-to-image node; the README states that it accepts up to 16 input or reference images through a ComfyUI batch.

Kie/Api · 4 inputs · 2 parameters · 2 outputs

KIE_GPTImage2_TextToImage

GPT Image text-to-image node in the documented GPT Image 2 and 2.5 family; the README describes model-aware aspect ratio and resolution controls.

Kie/Api · 3 inputs · 2 parameters · 2 outputs

KIE_GridSlice

Slice a grid image into equal tiles, with optional border cropping and gutter removal. - IMAGE: Grid image batch (BHWC float32 0–1). - grid: Layout (2x2 or 3x3). - outercroppx: Pixels trimmed from each side before slicing. - gutterpx: Pixels removed between tiles inside the grid. - order: row-major (top-left → right, then down) or column-major (top-left → down, then right). - processbatch: first to slice only the first image in the batch, all to slice every image and concatenate tiles. - log: Enable console logging. - IMAGE: Tile batch (4 tiles for 2x2, 9 tiles for 3x3 per processed image), BHWC float32 in [0, 1]. - Applies outer crop, then removes gutters, then slices into tiles. - Uses floor sizing when dimensions are not perfectly divisible; extra pixels on the right/bottom are dropped. - Validates crop/gutter to avoid zero or negative tile sizes; raises readable errors when impossible. - Ordering controls the tile sequence in the output batch. For example, row-major on 2x2 produces tiles in this order: 1. Top-left 2. Top-right 3. Bottom-left 4. Bottom-right

Kie/Helpers · 1 input · 6 parameters · 1 output

KIE_GrokImagine_I2I

Generate one or more images from a source image using Grok Imagine. - images (IMAGE, required): Source image batch. Only the first image is uploaded. - prompt (STRING, optional): Edit/style prompt text. Max length: 390000. - log (BOOLEAN, optional): Enable helper logging (default: true). - image (IMAGE): ComfyUI image batch (BHWC float32, 0..1). - taskid (STRING): KIE task id from createTask, for chaining into Grok I2V. - If multiple input images are connected, the node uploads only the first image and logs that choice. - The node downloads all returned resultUrls and returns them as one IMAGE batch. - This is the intended upstream source for Grok I2V taskid + index chaining.

Kie/Api · 1 input · 2 parameters · 2 outputs

KIE_GrokImagine_I2V

Generate a short video clip from either one uploaded external image or a Grok-generated image reference. - prompt Optional motion prompt. Maximum length: 5000 characters. - images Optional external source image batch. If connected, only the first image is uploaded and used. - taskid Optional Grok image-generation task id. Use this instead of images when you want to animate one of Grok's generated images. - index Which image to use from the Grok taskid source. Range: 0 to 5. - mode Choose from fun, normal, or spicy. Spicy is only supported for taskid sources. - duration Output length in seconds: 6, 10, or 15. - resolution Choose 480p or 720p. - log Enable console logging. - VIDEO ComfyUI VIDEO output compatible with the SaveVideo node. - External image mode Connect images and leave taskid empty. - Grok image reference mode Set taskid and index, and leave images disconnected. - Provide exactly one source method: images or taskid. - External image mode supports one image only. - If multiple ComfyUI images are connected, the first image is used. - If mode=spicy is selected with an external image, the node logs a warning and sends normal. - The KIE callback example for this endpoint appears to show a .jpg URL in resultUrls; this node treats the endpoint as video and downloads the first result as VIDEO output.

Kie/Api · 1 input · 7 parameters · 1 output

KIE_GrokImagine_T2I

Generate one or more images from a text prompt using Grok Imagine. - prompt (STRING, required): Generation prompt text. Max length: 5000. - aspectratio (COMBO, optional): 2:3, 3:2, 1:1, 9:16, 16:9 (default: 1:1). - log (BOOLEAN, optional): Enable helper logging (default: true). - image (IMAGE): ComfyUI image batch (BHWC float32, 0..1). - taskid (STRING): KIE task id from createTask, for chaining into Grok I2V. - The node downloads all returned resultUrls and returns them as one IMAGE batch. - This is the intended upstream source for Grok I2V taskid + index chaining. - Polling and timeout are handled internally by shared helpers.

Kie/Api · 0 inputs · 3 parameters · 2 outputs

KIE_GrokImagine_T2V

Generate a short video clip from a text prompt using the Grok Imagine text-to-video model. - prompt Text prompt describing the desired video motion. Maximum length: 5000 characters. - aspectratio Choose from 2:3, 3:2, 1:1, 9:16, or 16:9. - mode Choose from fun, normal, or spicy. - duration Output length in seconds: 6, 10, or 15. - resolution Choose 480p or 720p. - log Enable console logging. - VIDEO ComfyUI VIDEO output compatible with the SaveVideo node. Prompt: "A couple of doors open to the right one by one randomly and stay open, revealing tiny rooms with little people living inside." Mode: normal Aspect ratio: 16:9 Duration: 6 Resolution: 480p - All enum inputs are validated before the task is submitted. - This node uses internal defaults for polling, retries, and timeout handling. - The KIE callback example for this endpoint appears to show a .jpg URL in resultUrls; this node treats the endpoint as video and downloads the first result as VIDEO output.

Kie/Api · 0 inputs · 6 parameters · 1 output

KIE_Kling25_I2V_Pro

Generate a short video clip from a prompt and a required first image, with an optional tail image, using the Kling 2.5 Turbo I2V Pro model. - firstframe (IMAGE, required): Source image batch; the first image is used. - prompt (STRING, multiline, required): The text prompt for the video. - lastframe (IMAGE, optional): Optional tail frame; the first image is used if provided. - negativeprompt (STRING, multiline, optional): Optional negative prompt. - duration (COMBO, seconds, default: 5): Video length. Options: 5, 10. - cfgscale (FLOAT, default: 0.5): Guidance scale from 0.0 to 1.0. - log (BOOLEAN, default: true): Enable KIE console logs. - video (VIDEO): ComfyUI video output referencing a temporary .mp4 file. - If a job stalls or fails, check https://kie.ai/logs. - This node uses internal defaults for polling/retries/timeouts (not exposed in the UI).

Kie/Api · 2 inputs · 5 parameters · 1 output

KIE_Kling26_I2V

Generates video from a single input image using the Kling 2.6 image-to-video model.

Kie/Api · 1 input · 4 parameters · 1 output

KIE_Kling26_T2V

Generates video directly from a text prompt, with aspect ratio, duration, and sound options described by the API.

Kie/Api · 0 inputs · 5 parameters · 1 output

KIE_Kling26Motion_I2V

Image-to-video generation with additional motion-control parameters for more directed camera and motion behavior.

Kie/Api · 2 inputs · 4 parameters · 1 output

KIE_Kling3_Preflight

Validates and uploads inputs, returns the createTask payload JSON without running generation, and outputs kling_data for direct chaining into Kling 3.0 Video.

Kie/Helpers · 4 inputs · 8 parameters · 3 outputs

KIE_Kling3_Video

Experimental Kling 3.0 video node supporting single-shot or multi-shot generation, with optional first/last frame control and named element references.

Kie/Api · 5 inputs · 8 parameters · 1 output

KIE_Kling3Motion_I2V

Generates video from a required reference image and motion-driver video, with an optional prompt and ComfyUI VIDEO output.

Kie/Api · 2 inputs · 4 parameters · 1 output

KIE_KlingElements

Build one Kling element payload from images or video. - name (STRING): element name used in prompts with @name // super important must be elementname so elementdog then use @elementdog in the prompt - description (STRING, optional) - images (IMAGE, optional): image element source (2-4 images) - video (VIDEO, optional): video element source (single clip) - log (BOOLEAN, optional) - Provide exactly one media type: - images only, or - video only - Using both images and video in one element raises an error. - KIEELEMENT: single element payload - STRING: JSON preview of the element payload

Kie/Helpers · 2 inputs · 3 parameters · 2 outputs

KIE_KlingElementsBatch

Batch up to 9 KIEELEMENT payloads into one KIEELEMENTS payload for the Kling 3.0 node. - element1 to element9 (KIEELEMENT, optional) - At least one element is required. - Duplicate element names are rejected. - KIEELEMENTS: batched elements payload - STRING: JSON preview of batched elements

Kie/Helpers · 9 inputs · 0 parameters · 2 outputs

KIE_NanoBanana2_Image

Generate images using KIE's Nano Banana 2 model. This node supports optional reference images, Google web-search grounding, and the same async task flow as the other KIE image nodes. Maximum reference image inputs: 14. - Prompt Text description of the image to generate. - Images (optional) Up to 14 reference images used for guidance. - Google Search Enable Google web-search grounding for real-time information. - Aspect Ratio Output image aspect ratio. - Resolution 1K, 2K, or 4K. - Output Format jpg or png. - Poll Interval How often the node checks task status (seconds). - Timeout Maximum wait time before failing (seconds). - Log Enable progress output in the console. - IMAGE ComfyUI image tensor (BHWC, float32, range 0-1). - Credits are consumed per request. - If no images are connected, imageinput is sent as an empty list.

Kie/Api · 1 input · 6 parameters · 1 output

KIE_NanoBananaPro_Image

Generate high-quality images using KIE’s Nano Banana Pro model. This node supports optional reference images and exposes major generation controls directly in ComfyUI. Maximum reference image inputs: 8. - Prompt Text description of the image to generate. - Images (optional) Up to 8 reference images used for guidance. - Aspect Ratio Output image aspect ratio (e.g. 1:1, 16:9). - Resolution 1K, 2K, or 4K output resolution. - Output Format PNG or JPG. - Poll Interval How often the node checks task status (seconds). - Timeout Maximum wait time before failing (seconds). - Log Enable progress output in the console. - IMAGE ComfyUI image tensor (BHWC, float32, range 0–1). - Credits are consumed per request. - If the API is under heavy load, generation may take longer. - If no images are connected, imageinput is sent as an empty list. You can always visit https://kie.ai/logs to see the progress, why something may of failed. If their is a timeout you can usually see your final image/video here as well

Kie/Api · 1 input · 5 parameters · 1 output

KIE_Seedance15Pro_I2V

Higher-quality image-to-video generation using the Seedance 1.5 Pro model.

Kie/Api · 1 input · 7 parameters · 1 output

KIE_Seedance2_Preflight

Validates Seedance 2.0 payload structure and shows resolved media-field ordering before spending credits.

Kie/Helpers · 5 inputs · 9 parameters · 3 outputs

KIE_Seedance2_Video

Experimental video node supporting text-to-video, first-frame and first/last-frame image-to-video, multimodal references, and mixed frame/reference payloads.

Kie/Api · 6 inputs · 9 parameters · 1 output

KIE_SeedanceV1Pro_Fast_I2V

Convert a single reference image plus prompt into a short video clip using ByteDance’s rapid Seedance V1 Pro image-to-video model. - Prompt Describe how the source image should animate. - Images ComfyUI image tensor batch. Only the first frame is uploaded. - Resolution Choose 720p for faster jobs or 1080p for higher quality. - Duration Output length (5s or 10s). - Log Enable console logging for troubleshooting. - VIDEO ComfyUI VIDEO output compatible with the SaveVideo node. - The reference image is uploaded temporarily to the KIE API. - Credits are consumed per request. Monitor remaining credits with the provided node. - Video rendering can take a minute or more depending on resolution and duration. - This node uses internal defaults for polling/retries/timeouts (not exposed in the UI). Visit https://kie.ai/logs to monitor job progress or retrieve outputs if the node times out.

Kie/Api · 1 input · 4 parameters · 1 output

KIE_Seedream45_Edit

Edit an input image (or batch of images) using Seedream 4.5. - prompt (STRING, required): Edit prompt text (max 3000 chars). - images (IMAGE, required): Source image batch. Up to 14 images are uploaded; extra images are ignored. - aspectratio (COMBO, optional): 1:1, 4:3, 3:4, 16:9, 9:16, 2:3, 3:2, 21:9 (default: 1:1). - quality (COMBO, optional): basic, high (default: basic). - log (BOOLEAN, optional): Enable helper logging (default: true). - IMAGE: ComfyUI image tensor (BHWC float32, 0..1). - Uploads up to 14 images and sends all uploaded URLs in input.imageurls. - If more than 14 images are provided, only the first 14 are used and a log message is emitted.

Kie/Api · 1 input · 4 parameters · 1 output

KIE_Seedream45_TextToImage

Generate an image from a text prompt using Seedream 4.5. - prompt (STRING, required): Generation prompt text. - aspectratio (COMBO, optional): 1:1, 4:3, 3:4, 16:9, 9:16, 2:3, 3:2, 21:9 (default: 1:1). - quality (COMBO, optional): basic, high (default: basic). - log (BOOLEAN, optional): Enable helper logging (default: true). - image (IMAGE): ComfyUI image tensor (BHWC float32, 0..1). - This node is text-to-image only (no image input required). - Polling and timeout are handled internally by shared helpers.

Kie/Api · 0 inputs · 4 parameters · 1 output

KIE_Suno_Music_Advanced

Create a Suno music generation task via KIE API. This node exposes advanced weights and returns two AUDIO outputs + two cover images. - title (STRING, required): Track title (required in custom mode). - style (STRING, required): Style text (required in custom mode, multiline). - prompt (STRING, required): Prompt text (lyrics in custom mode when instrumental is false). - custommode (BOOLEAN, required): Enable custom mode. - instrumental (BOOLEAN, required): Instrumental-only mode. - model (COMBO, required): V4, V45, V45PLUS, V45ALL, V5 Optional: - negativetags (STRING): Optional. - vocalgender (COMBO): male or female (custom mode only). - styleweight (FLOAT): 0..1 - weirdnessconstraint (FLOAT): 0..1 - audioweight (FLOAT): 0..1 - log (BOOLEAN): Enable console logging. - audio1 (AUDIO): Generated audio 1. - audio2 (AUDIO): Generated audio 2. - data (STRING): Full API response JSON (formatted). - image1 (IMAGE): Generated cover image 1. - image2 (IMAGE): Generated cover image 2. - Non-custom mode expects prompt only (500 chars max); other fields should be empty. - Custom mode requirements: - Instrumental: style + title - Non-instrumental: style + title + prompt - Polling parses both callback-style (data.data[].audiourl) and record-info style (data.response.sunoData[].audioUrl) responses.

Kie/Api · 0 inputs · 12 parameters · 5 outputs

KIE_Suno_Music_Basic

Create a Suno music generation task via KIE API. This node keeps inputs minimal and returns two AUDIO outputs + two cover images. - title (STRING, required): Track title (required in custom mode). - style (STRING, required): Style text (required in custom mode, multiline). - prompt (STRING, required): Prompt text (lyrics in custom mode when instrumental is false). - custommode (BOOLEAN, required): Enable custom mode. - instrumental (BOOLEAN, required): Instrumental-only mode. - model (COMBO, required): V4, V45, V45PLUS, V45ALL, V5 Optional: - negativetags (STRING): Optional. - vocalgender (COMBO): male or female (custom mode only). - log (BOOLEAN): Enable console logging. - audio1 (AUDIO): Generated audio 1. - audio2 (AUDIO): Generated audio 2. - data (STRING): Full API response JSON (formatted). - image1 (IMAGE): Generated cover image 1. - image2 (IMAGE): Generated cover image 2. - Non-custom mode expects prompt only (500 chars max); other fields should be empty. - Custom mode requirements: - Instrumental: style + title - Non-instrumental: style + title + prompt

Kie/Api · 0 inputs · 9 parameters · 5 outputs

KIE_SystemPrompt_Selector

Combine a user prompt with a system prompt template stored in prompts/images or prompts/videos. - userprompt (STRING, required): The user prompt text. - systemtemplate (COMBO, required): Template name pulled from prompts/images/.txt and prompts/videos/.txt. - STRING: Combined prompt (system + user). Each .txt file in prompts/images/ or prompts/videos/ must include: name: system prompt below - If {userprompt} is present, it is replaced with the user prompt. - If not present, the user prompt is appended to the end of the template. - Files missing name: or system prompt below are ignored.

Kie/Helpers · 0 inputs · 2 parameters · 1 output

KIEParsePromptGridJSON

Parses structured JSON output, such as from an LLM, into individual prompt outputs for multi-image and storyboard-style workflows.

Kie/Helpers · 1 input · 4 parameters · 12 outputs