Queue a ComfyDeploy run, poll for results, download images, and return them as a ComfyUI IMAGE tensor—using the SAME image handling patterns as the FAL example node (tensor->dataURI for input, PIL->np->tensor for output).

PVL/API · 1 input · 7 parameters · 1 output

PVL_ComfyDeploy_LoraPrep

Single-image in → single-image out ComfyDeploy node. - Exactly one IMAGE input (named `image`). - Exactly one IMAGE output ([1, H, W, 3]). - No batch parameter (internally enforces batch=1 on the API). - Removed text inputs (style, prompt).

PVL/API · 1 input · 5 parameters · 1 output

PVL_ComfyDeploy_Universal

Flexible ComfyDeploy node with LoraPrep-style image handling: - 10 (parameter_name_i, value_i) pairs. Values are OPTIONAL and ANY type. - If a value is an IMAGE tensor and you set the corresponding name to 'input_image', it will be serialized to a data URI exactly like your LoraPrep node. - Set upload_tensors=True to try uploading PNG bytes and pass a URL instead. - Collects all produced image URLs and returns a single IMAGE batch [B,H,W,3].

PVL/API · 1 output

PVL_Crop2AR

Compact description unavailable for this node.

PVL Tools · 1 input · 1 parameter · 1 output

PVL_CropToMask

Compact description unavailable for this node.

PVL Nodes/Image · 2 inputs · 3 parameters · 2 outputs

PVL_EdgeExtend

Edge extension node with iterative Nuke-style propagation. Pass logic: • premult_after_levels = (unprem * adjusted_alpha) • Working state per iteration: work_rgb_premult (RGB premult stream), work_alpha (coverage for UNDER) • For each enabled pass with radius r: 1) blur work_rgb_premult & work_alpha → unprem_blurred 2) band_out = clamp(dilate(work_alpha, r) - work_alpha, 0, 1) 3) work_rgb_premult = work_rgb_premult + (unprem_blurred * band_out)[premult] * (1 - work_alpha) work_alpha = work_alpha + band_out * (1 - work_alpha) 4) (restore detail) hi-freq add-back limited to intermediate alpha & inner edge band • Diagnostics after pass n: unprem(work_rgb_premult, work_alpha) • Final premult for output uses ORIGINAL alpha × opacity (unchanged), or switchable later if desired. • Optional linear workflow (work_in_linear)

Image/Alpha · 2 inputs · 4 parameters · 7 outputs

PVL_fal_DepthAnythingV2_API

ComfyUI node for FAL 'fal-ai/image-preprocessors/depth-anything/v2' — Depth Map Estimation. Inputs: - image (IMAGE): One input image (batch NOT supported). Outputs: - IMAGE: Depth map image tensor (B,H,W,C).

PVL Tools · 1 input · 0 parameters · 1 output

PVL_fal_EvfSam_API

Compact description unavailable for this node.

PVL Tools FAL · 1 input · 11 parameters · 1 output

PVL_fal_EvfSam_X5_API

Compact description unavailable for this node.

PVL Tools FAL · 4 inputs · 14 parameters · 5 outputs

PVL_fal_Flux_Dev_Inpaint_API

PVL Flux Dev Inpaint (fal.ai) Endpoint: fal-ai/flux-lora/inpainting Inputs: - image: ComfyUI IMAGE - mask: ComfyUI MASK (float 0..1, where 1.0 = inpaint region) Implementation: - Uses Base64 PNG data URIs for image_url + mask_url (no fal.storage upload). - Queue submit -> poll -> fetch (mirrors PVL Flux With LoRA node pattern).

PVL Tools FAL · 2 inputs · 15 parameters · 1 output

PVL_fal_Flux2_Camera_Ctrl_API

Compact description unavailable for this node.

PVL Tools FAL · 0 inputs · 12 parameters · 1 output

PVL_fal_Flux2_Dev_API

Compact description unavailable for this node.

PVL Tools FAL · 7 inputs · 20 parameters · 1 output

PVL_fal_Flux2_Flex_API

Compact description unavailable for this node.

PVL Tools FAL · 7 inputs · 14 parameters · 1 output

PVL_fal_Flux2_Pro_API

Compact description unavailable for this node.

PVL Tools FAL · 7 inputs · 11 parameters · 1 output

PVL_fal_FluxDev_API

Compact description unavailable for this node.

PVL Tools FAL · 0 inputs · 11 parameters · 1 output

PVL_fal_FluxDevPulidAvatar_API

Compact description unavailable for this node.

PVL Tools FAL · 1 input · 5 parameters · 1 output

PVL_fal_FluxGeneral_API

Compact description unavailable for this node.

PVL Tools FAL · 1 input · 28 parameters · 1 output

PVL_fal_FluxPro_Fill_API

Compact description unavailable for this node.

PVL Tools · 2 inputs · 6 parameters · 1 output

PVL_fal_FluxPro_v1_1_Ultra_API

Compact description unavailable for this node.

PVL Tools · 1 input · 11 parameters · 1 output

PVL_fal_FluxPulid

ComfyUI node for fal-ai/flux-pulid endpoint. Generates an image using a prompt and optional reference image.

PVL/FAL · 1 input · 12 parameters · 2 outputs

PVL_fal_FluxWithLora_API

Compact description unavailable for this node.

PVL Tools FAL · 0 inputs · 19 parameters · 1 output

PVL_fal_FluxWithLoraPulID_API

Compact description unavailable for this node.

PVL Tools FAL · 1 input · 20 parameters · 1 output

PVL_fal_Kontext_Dev_API

Compact description unavailable for this node.

PVL Tools · 1 input · 10 parameters · 1 output

PVL_fal_KontextDevInpaint_API

Compact description unavailable for this node.

PVL Tools FAL · 3 inputs · 16 parameters · 1 output

PVL_fal_KontextDevLora_API

Compact description unavailable for this node.

PVL Tools · 1 input · 12 parameters · 1 output

PVL_fal_KontextMaxMulti_API

Compact description unavailable for this node.

PVL Tools FAL · 4 inputs · 9 parameters · 1 output

PVL_fal_KontextMaxSingle_API

Compact description unavailable for this node.

PVL Tools FAL · 1 input · 8 parameters · 1 output

PVL_fal_KontextPro_API

Compact description unavailable for this node.

PVL Tools · 1 input · 8 parameters · 1 output

PVL_fal_LumaPhoton_FlashReframe_API

Compact description unavailable for this node.

PVL Tools · 1 input · 8 parameters · 1 output

PVL_fal_LumaPhoton_Reframe_API

Compact description unavailable for this node.

PVL Tools · 1 input · 8 parameters · 1 output

PVL_fal_Moondream3_Segment_API

ComfyUI node for FAL 'fal-ai/moondream3-preview/segment' — Segmentation (Moondream 3 preview). - Uses Base64 Data URI for image_url (no upload) - TRUE PARALLEL batch: submit all -> poll all -> pack results - Supports: object (string), spatial_references (JSON), preview (bool), settings {temperature, top_p, max_tokens} Returns: mask_preview: IMAGE (mask image as RGB for easy viewing; black if not returned) mask: MASK (mask alpha or luminance; zeros if not returned) path: STRING (SVG path data; JSON array aligned to outputs) bbox_json: STRING (bbox object; JSON array aligned to outputs) finish_json: STRING (finish_reason; JSON array aligned to outputs) usage_json: STRING (usage_info; JSON array aligned to outputs) spatial_references_json format (normalized coords 0..1), examples: - Points: [{"x": 0.64, "y": 0.40}] - Arrays of 2 floats (x,y): [[0.64, 0.40]] - Arrays of 4 floats (x1,y1,x2,y2): [[0.10, 0.20, 0.40, 0.50]]

PVL Tools · 1 input · 7 parameters · 6 outputs

PVL_fal_NanoBanana_API

ComfyUI node for FAL 'fal-ai/nano-banana/edit' — batch-friendly image edit using a prompt. Inputs: - prompt (STRING): Edit instruction for the model. - image (IMAGE): One or more input images (batch supported). - num_images (INT): Number of output images to generate. - output_format (CHOICE): "jpeg" or "png" (API default is "jpeg"). - sync_mode (BOOLEAN): If true, API can return data-URIs instead of hosted URLs. Outputs: - IMAGE: Edited image tensor (B, H, W, C) in [0, 1]. - STRING: Optional text/description returned by the endpoint, if any.

PVL Tools · 1 input · 4 parameters · 2 outputs

PVL_fal_QwenImage_API

Compact description unavailable for this node.

PVL Tools FAL · 0 inputs · 12 parameters · 1 output

PVL_fal_RemoveBackground_API

ComfyUI node for FAL 'fal-ai/birefnet/v2' — Remove Background V2. Features: - TRUE PARALLEL execution: submits all requests first, then polls all in parallel - Error handling for individual requests with partial results support - Batch processing with worker cap to prevent thread explosion - Optional sync_mode toggle Inputs: - image (IMAGE): Input image(s). Batch processing is supported with parallel API calls. - model (CHOICE): Which model variant to use. - operating_resolution (CHOICE): Resolution for inference ("1024x1024" or "2048x2048"). - output_format (CHOICE): "png" or "webp". - output_mask (BOOLEAN): Whether to also return the mask. - refine_foreground (BOOLEAN): Whether to refine the foreground (default: True). - sync_mode (BOOLEAN): Use synchronous mode for FAL API (default: False). Outputs: - IMAGE: Foreground with background removed (RGB or RGBA if provided). - MASK: Optional mask (1-channel, float, shape [B,H,W]) if output_mask=True.

PVL Tools · 1 input · 6 parameters · 2 outputs

PVL_fal_Sam3_Segmentation_API

ComfyUI node for FAL 'fal-ai/sam-3/image' — Segment Image (SAM-3). Key features: - Base64 Data URI submission (no storage upload) - TRUE PARALLEL batch execution: Phase 1: Submit all -> get request_ids immediately Phase 2: Poll all -> fetch results in parallel - Supports prompt, point prompts, box prompts, apply_mask, output_format, sync_mode - Returns: preview: IMAGE (the API's primary preview image) mask: MASK (the FIRST returned mask per input image; if none, zeros) scores_json: STRING (JSON array aligned to input batch) boxes_json: STRING (JSON array aligned to input batch) Notes: - The API can return multiple masks. This node returns only the *first* mask as MASK output (because ComfyUI MASK output is typically one mask per input image). The other masks' metadata (scores/boxes) are included in JSON outputs. - point_prompts_json and box_prompts_json are optional JSON strings: point_prompts_json example: [{"x": 120, "y": 220, "label": 1, "object_id": 0}, {"x": 10, "y": 20, "label": 0}] box_prompts_json example: [{"x_min": 10, "y_min": 20, "x_max": 400, "y_max": 500, "object_id": 0}]

PVL Tools · 1 input · 9 parameters · 4 outputs

PVL_fal_SeeDream_45_API

PVL SeeDream 4.5 (fal.ai) - If any image input is connected -> use edit endpoint: "fal-ai/bytedance/seedream/v4.5/edit" and send images as Base64 data URIs via image_urls. - If no image is connected -> use text-to-image endpoint: "fal-ai/bytedance/seedream/v4.5/text-to-image"

PVL Tools FAL · 1 output

PVL_fal_SeeDream4_API

ComfyUI node for FAL 'fal-ai/bytedance/seedream/v4/edit' — image editing using prompt and uploaded input images. Inputs: - prompt (STRING): Edit instruction for the model. - image (IMAGE): One or more input images (batch supported). - image_size (CHOICE): Only used for UI preset, width & height are always passed to API. - width / height (INT): Pixel dimensions passed to the API. - seed (INT): Optional seed for deterministic results. - num_images (INT): Number of outputs to generate. - sync_mode (BOOLEAN): Whether to wait for results directly (may increase latency). Outputs: - IMAGE: Edited image tensor (B, H, W, C) in [0, 1]. - STRING: Optional text/description returned by the endpoint, if any.

PVL Tools · 1 input · 7 parameters · 2 outputs

PVL_Gemini_API

Compact description unavailable for this node.

PVL/LLM · 2 inputs · 13 parameters · 1 output

PVL_Gemini_API_avatar_special

PVL Gemini API (Avatar Special) Adds retry on timeout and linear backoff. Optional OpenAI fallback / override for failed Gemini calls.

PVL/LLM · 3 inputs · 17 parameters · 1 output

PVL_Gemini_API_Multi

Compact description unavailable for this node.

PVL/LLM · 7 inputs · 13 parameters · 1 output

PVL_Gemini_with_fallback_API

Compact description unavailable for this node.

PVL/LLM · 3 inputs · 16 parameters · 1 output

PVL_Gemini_with_fallback_API_Multi

Compact description unavailable for this node.

PVL/LLM · 9 inputs · 15 parameters · 1 output

PVL_GeminiSegmentation

Multi-image segmentation using Gemini 2.5 via REST.

Image/Segmentation · 0 inputs · 13 parameters · 10 outputs

PVL_Get_Image_Size

PVL Tools/Image · 1 input · 0 parameters · 6 outputs

PVL_Google_NanoBanana_API

Compact description unavailable for this node.

1 input · 19 parameters · 2 outputs

PVL_Google_NanoBanana_API_mandatory_IMG

Compact description unavailable for this node.

1 input · 19 parameters · 2 outputs

PVL_Google_NanoBanana_Multi_API

Compact description unavailable for this node.

8 inputs · 19 parameters · 2 outputs

PVL_Google_NanoBanana_PRO_Multi_API

Compact description unavailable for this node.

8 inputs · 21 parameters · 2 outputs

PVL_Image_Translation_Offset

Compact description unavailable for this node.

Image/Alignment · 3 inputs · 3 parameters · 2 outputs

PVL_ImageBatchToImageList

Compact description unavailable for this node.

PVL/Utils/Image · 1 input · 0 parameters · 1 output

PVL_ImageComposite

Compact description unavailable for this node.

Pvl/Image Manipulation · 3 inputs · 5 parameters · 1 output

PVL_ImageListToImageBatch

Compact description unavailable for this node.

PVL/Utils/Image · 1 input · 0 parameters · 1 output

PVL_ImageResize

PVL Tools · 3 outputs

PVL_ImageStitch

PVL – Image Stitch: stitch up to 8 images either linearly or via pack-to-square.

Image/Compose · 8 inputs · 9 parameters · 1 output

PVL_Math

Compact description unavailable for this node.

Essentials Mb/Utilities · 2 outputs

PVL_MathCondition

Compact description unavailable for this node.

Essentials Mb/Utilities · 2 outputs

PVL_NoneOutputNode

A node that can either pass through its input or output None to skip downstream execution. Input is optional - works even when nothing is connected.

Utility · 1 output

PVL_Padding

Compact description unavailable for this node.

PVL/Utility · 1 input · 8 parameters · 1 output

PVL_Qwen_Img_Edit_Inpaint

ComfyUI node for FAL 'fal-ai/qwen-image-edit/inpaint' — image editing with mask & prompt. Inputs: - prompt (STRING): Edit instruction for the model. - image (IMAGE): Input image tensor (batch or single). - mask (IMAGE, optional): Mask image for inpainting (white=edit, black=keep). - negative_prompt (STRING, optional): Negative prompt for generation. - num_inference_steps (INT): Diffusion steps (default: 30). - guidance_scale (FLOAT): CFG scale (default: 4.0). - num_images (INT): Number of output images to generate. - enable_safety_checker (BOOLEAN): Enable safety filtering (default: True). - output_format (CHOICE): jpeg / png (default: png). - acceleration (CHOICE): none / regular / high (default: regular). - strength (FLOAT): Inpainting noise strength (0–1, default: 0.93). - sync_mode (BOOLEAN): Whether to request data URIs instead of URLs (default: False). - seed (INT): Optional random seed (default: 0 -> random). Outputs: - IMAGE: Generated edited image tensor (B, H, W, C) in [0, 1]. - STRING: Prompt used (for reference).

PVL Tools · 2 inputs · 11 parameters · 2 outputs

PVL_Qwen_Img_Edit_Plus

ComfyUI node for FAL 'fal-ai/qwen-image-edit-plus' — multi-image edit using parallel API calls. Features: • Parallel API calls (num_images determines number of requests) • All input images submitted to each call (combined context) • Retry with linear backoff (1–2–3 s) • Timeout control • Aspect ratio selector ('custom' + presets) • Safety checker toggle (default: off) • Negative prompt support (sent to API) • Delimiter-based prompt splitting • Debug logging • Seed auto-increment per call: seed, seed+1, …

PVL Tools · 0 inputs · 1 parameter · 2 outputs

PVL_SaveOrNot

Compact description unavailable for this node.

PVL Tools · 1 input · 4 parameters · 0 outputs

PVL_Seg_Florence2_API

Compact description unavailable for this node.

PVL Tools FAL · 5 inputs · 10 parameters · 10 outputs

PVL_SplitString

Compact description unavailable for this node.

PVL Utils · 0 inputs · 2 parameters · 6 outputs

PVL_Stitch2Size

Stitches up to 10 images into a single canvas (width x height), choosing an optimal grid, scaling proportionally, and padding with a chosen color. If keep_relative_scale=True, a single global scale factor is used for all images (tightest fit across the grid), preserving their relative sizes. Otherwise, each image scales independently to best fit its own cell.

PVL/Image · 0 inputs · 4 parameters · 1 output

PVL_StringConcatX3

PVL String Concat x3 Takes three string inputs and returns their direct concatenation. Preserves all whitespace exactly as provided (no extra spaces added).

🧩 PVL/Strings · 0 inputs · 3 parameters · 1 output

PVL_StringConcatX8

PVL String Concat x8 Takes eight string inputs and returns their direct concatenation. Preserves all whitespace exactly as provided (no extra spaces added).

🧩 PVL/Strings · 0 inputs · 8 parameters · 1 output

PVL_StringListToString

Compact description unavailable for this node.

PVL/Utils/Text · 0 inputs · 2 parameters · 1 output

PVL_StringToNumber

Compact description unavailable for this node.

PVL Utils · 1 input · 1 parameter · 2 outputs

PVL_StringToStringList

Compact description unavailable for this node.

PVL/Utils/Text · 0 inputs · 2 parameters · 1 output

PVL_StylePicker

Compact description unavailable for this node.

PVL/Style · 2 outputs

PVL_Switch

Choose between two inputs based on a boolean switch, evaluating only the selected branch.

🟣PVL/Utility · 1 input · 1 parameter · 1 output

PVL_Switch_Huge

Compact description unavailable for this node.

PVL Tools · 11 inputs · 11 parameters · 1 output

PVL_Switch_x10

Compact description unavailable for this node.

PVL Tools · 0 inputs · 1 parameter · 1 output

PVL_Switch_x15

Compact description unavailable for this node.

PVL Tools · 0 inputs · 1 parameter · 1 output

PVL_Switch_x5

Compact description unavailable for this node.

PVL Tools · 0 inputs · 1 parameter · 1 output

PVL_Text_Overlay

⚙️PVL Nodes/Text · 1 output

PVL_Txt

Compact description unavailable for this node.

PVL · 0 inputs · 1 parameter · 1 output

PVLCheckIfConnected

Compact description unavailable for this node.