Directorio de recursos
Encuentra lo que necesita tu flujo.
Guides: techniques, setup, and troubleshootingBusca Pipes de Sweet Tea, flujos externos, modelos, paquetes de nodos y guías en un catálogo con evidencias.
Empieza con un flujo ↗Resultados de búsqueda
Todos los recursos
20 mostrados of 41.727
Topaz creative image enhancement
Enhance and resize an uploaded image with Topaz Reimagine, returning one processed image. For a first run, upload your source image and optionally enter an Enhancement prompt describing details or an aesthetic to emphasize or retain. Leave the prompt empty for enhancement without text guidance. The defaults request 1080 × 774 pixels with Creativity 3, Enhance faces and Preserve facial identity enabled, and Preserve colors disabled. Enter output dimensions in pixels, using values from 0 to 32000 even though the dimension input widgets accept a wider integer range. Set one dimension to 0 for automatic sizing from the other, or both to 0 to use the original size. When the requested aspect ratio differs from the source, the default fit letterboxes the image. Enable Crop to fill to fill the requested frame by cropping. Creativity ranges from 1 to 9. Subject detection offers All, Foreground, or Background. Enable Enhance faces to apply face treatment; Face creativity and Face sharpness each range from 0 to 1 and apply to that treatment. Face sharpness controls how sharp enhanced faces are relative to the background. Preserve facial identity and Preserve colors separately request retention of likeness and the original palette. The model is fixed to Reimagine. Creative enhancement can change image details and should not be treated as faithful historical reconstruction. There is no seed control. Uploaded transparency is not preserved through an alpha-mask path: only the loader's image output reaches enhancement. Processing sends the image to the remote Topaz partner API and requires configured API access and authentication.
Anima edge-guided illustration
Generate prompted Anima illustrations guided by an uploaded image’s outlines, and receive an edge-map preview alongside the generated images. For a first run, upload a reference with the composition you want and replace the example positive prompt with your desired subject, colors and style. Review the negative prompt too: its example includes sepia, which conflicts with the positive example’s sepia palette. Keep output size at 1 megapixel, control strength at 1, batch size at 1 and Turbo mode off. Seed -1 chooses fresh random noise each run; use a fixed seed when comparing settings. The reference is processed into dark edges on a light background. Its proportions determine the output proportions, subject to latent sizing. It supplies outline guidance rather than an encoded starting image, so uploaded colors, texture and transparency are not preserved. The edge-map preview is shown before resizing; the guide used for generation is resized to the requested megapixel target. Control strength adjusts outline influence; 0 removes that influence while retaining reference-derived sizing. Edge detection thresholds range from 0.01 to 0.99: keep the low threshold below the high threshold. Lower thresholds can retain more contours; higher thresholds can reduce clutter. Control start and end are fractions of sampling from 0 to 1; keep start below end to define the interval during which guidance applies. Guide interpolation controls how the edge map is resized. Turbo mode jointly selects the fixed Turbo LoRA and the separate Turbo steps and CFG settings. Standard defaults are 30 steps and CFG 4; Turbo defaults are 8 steps and CFG 1. Only the active mode’s steps and CFG affect sampling, and Turbo LoRA strength affects the selected model only when Turbo is enabled. Use positive step counts. Sampler and scheduler apply to both modes; the source presets are Euler and simple. Keep denoise at 1 initially: lowering it changes sampling from the empty latent and does not turn this into image-to-image preservation. Batch size controls the number of generated variants sharing the prompt and guide. Larger batches and output sizes increase memory requirements. The preview remains a guide preview rather than another generated variant. The fixed model setup uses anima-base-v1.0.safetensors in diffusion_models, anima-lllite-any-test-like-v2.safetensors in model_patches, qwen_3_06b_base.safetensors in text_encoders, qwen_image_vae.safetensors in vae, and anima-turbo-lora-v0.2.safetensors in loras. These files are referenced by the workflow; their availability is not established by this review. The base model, text encoder and VAE come from circlestone-labs/Anima on Hugging Face, the patch from Comfy-Org/Anima-LLLite, and the LoRA from circlestone-labs/Anima-Official-LoRAs. Generated-image transport also requires the declared tea-time-labs/sweet-tea-nodes package.
Vidu Q2 text to video
Generate a short video from a written scene description using Vidu Q2, with optional background music. No source image is required. Generation sends your prompt and settings to Vidu through a paid provider call. For a first run, use the supplied scene prompt with the default 5-second duration, 16:9 aspect ratio, and 720p resolution. Run the Pipe to return one generated video to Sweet Tea. To create your own scene, describe the subject, action, setting, camera movement, and visual style in a prompt of up to 2,000 characters. Duration ranges from 1 to 10 seconds in whole-second increments. Choose landscape framing with 16:9 or 4:3, portrait framing with 9:16 or 3:4, or square framing with 1:1. Resolution can be 720p or 1080p. Enable Background music to request music with the generated video. Seed -1 randomizes each run. A fixed seed supplies a specific value to the provider without guaranteeing identical results. Output format and codec default to automatic selection; the output stage returns the video to Sweet Tea rather than saving it through the original ComfyUI SaveVideo node.
Flux.2 Dev image edit
Generate edited images from a reference image and a written instruction using Flux.2 Dev. For a first run, upload a source image, describe the changes you want and the features to retain, and keep the supplied defaults: 1 megapixel, Lanczos resizing, one image, guidance 4, 20 sampling steps and Euler. The example instruction requests anime styling while retaining identity. Seed -1 selects random sampling noise. Output megapixels sets the target total pixel count of the reference while following its aspect ratio. The resized reference supplies editing conditioning and determines generation dimensions. Resize method changes reference interpolation before encoding; it does not add an output upscaling stage. Advanced controls provide image count, seed, Flux guidance, sampling steps and sampler choice. A fixed nonnegative seed reuses the noise setting; changing the reference, instruction or sampling settings can still change the result. Guidance adjusts instruction conditioning rather than the amount of source-image denoising. Generation starts from an empty latent, with the encoded source attached as reference conditioning. This tool has no text-only mode, negative-prompt control, image-to-image denoise-strength control or mask-based editing mode. Uploaded alpha transparency is not preserved by the connected RGB image path. The selected output contains generated images. The workflow uses flux2_dev_fp8mixed.safetensors, mistral_3_small_flux2_bf16.safetensors as its Flux.2 text encoder, and flux2-vae.safetensors.
Topaz creative upscale
Upload an image to produce one enhanced image at a scale-based size with Topaz Reimagine. For a first run, keep the 4× scale and creativity of 3; leave the enhancement prompt empty or describe details and appearance to guide the result. Scale multiplies the source width and height, approximately preserving its aspect ratio. Values below 1 request a smaller image. Keep each calculated dimension within 32,000 pixels. The sizing branch supplies dimensions; Topaz receives the original image directly. Processing sends the image to Topaz's external service. Creativity controls reinterpretation on a 1–9 scale; inspect details that must remain faithful to the source. Subject detection offers All, Foreground, and Background modes. Enhance faces enables additional face treatment; Face creativity and Face sharpness apply when it is enabled. Preserve face identity and Preserve colors request retention of facial identity and original colors. Crop to fill crops instead of letterboxing when output and source aspect ratios differ. Scale-based sizing normally preserves the source aspect ratio, so this setting usually has little effect. The model is fixed to Reimagine, and independent width and height overrides are unavailable. Uploaded transparency is not retained through the connected RGB image path.
Topaz video upscale
Enhance an uploaded video with Topaz Starlight (Astra) Fast and return one video at 1080p or 4K. For a first run, upload your source video, keep FullHD (1080p) output and Low compression, and leave the advanced export settings on auto. Run the Pipe to receive the enhanced video. Choose 4K (2160p) when that is your desired output resolution. Compression level selects the Topaz service's CQP compression setting: Low, Mid or High. Advanced export settings control the returned video's codec and container separately from the service's compression setting. Leave both on auto unless you need a particular format. Explicit codec values are h264 and av1; explicit container values are mp4, mkv and webm. Choose a compatible combination. Processing sends the source video to the authenticated Topaz provider API and uses provider-priced enhancement. Execution requires provider access. This Pipe uses the fixed Starlight (Astra) Fast model with frame interpolation disabled. It does not offer alternative enhancement models, creative prompts, frame-rate changes or slow motion.
Topaz video enhancement
Enhance an uploaded video to 1080p or 4K, optionally add frame interpolation, and receive one processed video. For a first run, upload your source video and keep Starlight (Astra) Fast, FullHD (1080p), interpolation Disabled, and compression Low. Leave the advanced output settings at their defaults. Configured Comfy partner API access is required: the video is sent to the Topaz partner API for provider-priced processing. Choose an enhancement model and target resolution to enhance the video. Resolution applies only when enhancement is enabled. Select apo-8 to enable interpolation; choose Disabled for the enhancement model if you want interpolation alone. Keep at least one operation enabled. Model-specific creativity, scene prompts, sharpness, and realism controls are not exposed. Interpolation frame rate and slow-motion controls are also not exposed; these settings rely on the node's schema defaults. Compression level controls the Topaz CQP processing setting. Advanced video output controls separately select the final container, codec, and CRF encoding quality. Available containers are auto, mp4, mkv, and webm; available codecs are auto, h264, and av1. Choose compatible container and codec settings for your playback software.
Seedance first and last frame video
Generate a prompted video transition between two uploaded images, with optional audio on Seedance 1.5 Pro. Both first and last frame images are required. This Pipe sends the images and prompt to ByteDance Seedance through a paid provider call and returns one generated video to Sweet Tea. For a first run, upload both images, replace the capybara example prompt with motion, transformation, and camera instructions suited to them, and keep the default Seedance 1.5 Pro model, 720p resolution, adaptive aspect ratio, and five-second duration. The example describes six shots and six seconds of audio; adapt its scope and timing to your requested clip. Duration sets the requested length from 3 to 12 seconds in whole-second increments. Resolution offers 480p, 720p, and 1080p. Aspect ratio offers adaptive sizing or a fixed frame proportion. Consider both endpoint compositions when choosing a fixed ratio. Model selection offers Seedance 1.5 Pro, 1.0 Pro, and 1.0 Lite. Generate audio is enabled by default and applies only to Seedance 1.5 Pro; describe desired music, sounds, or ambience in the prompt. Other available models ignore that control. Fix camera adds a provider instruction for a stationary camera, but the effect is not guaranteed. Avoid conflicting camera movement instructions when enabling it. Seed -1 requests a new random seed each run; a fixed seed controls generation randomization. Watermark optionally requests an AI-generated label. The image inputs use RGB image data; uploaded transparency is not carried through this workflow. The output is returned through Sweet Tea's video preview transport rather than the original local SaveVideo stage.
Z-Image-Turbo: Text to image
Generate images from a text prompt with the fixed Z-Image-Turbo model. No uploaded image is required. For a first run, replace the example prompt with a description of your subject, composition, setting, lighting and visual style. Keep 1024 × 1024 pixels, one image and eight sampling steps, then run the tool to receive the generated image. The workflow uses z_image_turbo_bf16.safetensors, qwen_3_4b.safetensors and ae.safetensors; these model files must be available to the runtime. Set width and height together to choose the aspect ratio. Both dimensions must be multiples of 16. Image count generates a batch from the same prompt. Larger dimensions and batches increase memory use and runtime. In advanced sampling settings, seed -1 selects a new random seed on each run. Use a fixed seed to compare changes under the same configuration. The source defaults are eight steps, guidance scale 1, res_multistep sampler and simple scheduler. Sampler, scheduler, steps and sampling shift affect the sampling process and interact; sampling shift defaults to 3. More steps do not guarantee a better image. Denoise amount defaults to 1. This tool starts from an empty latent, so reducing denoise changes the sampling range rather than preserving an existing image. Negative conditioning is zeroed internally, with no editable negative prompt. The model, text encoder and VAE are fixed; this tool provides text-to-image generation and batch generation, with no image-to-image mode.
Flux Fill outpaint
Generate an expanded image with prompt-guided surroundings using Flux Fill. For a first run, upload the image to extend and replace the example scene prompt with a description of the complete scene, lighting, style, and surroundings you want to add. Keep the other defaults and run the Pipe. It returns one image. The default canvas adds 400 pixels on the left, right, and bottom, with no extension above. Choose each extension in pixels, in increments of 8. Set a side to zero to leave that edge unextended. The padded canvas width is the source width plus the left and right extensions; its height is the source height plus the top and bottom extensions. Boundary feathering sets the width in pixels of the soft mask transition into the original image near extended edges. Advanced controls include Flux guidance for prompt influence and the separate sampler CFG, which defaults to 1. Sampling defaults are 20 steps, Euler, the normal scheduler, and denoise 1. Denoise controls sampling strength; lower values retain more of the initial latent and may leave added regions insufficiently filled. Seed -1 chooses a new random variation each run; use a fixed seed when comparing settings. Use sampling mask is disabled by default. Enable it to restrict sampling using the padding mask and its feathered boundary. This mode can substantially change results and may not work well with the model. Differential mask strength controls differential diffusion's mask handling when masked sampling is enabled, from 0 for no effect to 1 for the full effect. There is no final compositing step that restores the original pixels, so the original area can change; masked sampling does not promise exact pixel preservation. The mask comes from canvas padding, not uploaded transparency. This Pipe does not accept a painted inpainting mask or preserve uploaded alpha. Negative conditioning is automatically zeroed, so there is no negative-prompt field.
HappyHorse text to video
Create one video from an English or Chinese scene description using the HappyHorse API. Describe the subject, action, setting, camera movement and visual style; no uploaded media is required. For a first run, keep the supplied horse scene or write your own prompt, then use the default HappyHorse 1.0 model, 5-second duration, 16:9 aspect ratio and 720P resolution. Match the amount of action in your prompt to the chosen duration. Run the Pipe to receive one video. Choose HappyHorse 1.0 or 1.1. Both expose durations of 3–15 whole seconds, 720P or 1080P resolution, and aspect ratios of 16:9, 9:16, 1:1, 4:3 or 3:4. Model, resolution and duration affect provider generation cost. Generation sends your prompt to the provider and uses a paid API call. Seed -1 randomizes each run. A fixed seed supplies a specific generation seed without guaranteeing identical provider results. Enable Watermark to request an AI-generated watermark on the result. Advanced Video output settings offer automatic container and codec selection or explicit MP4, MKV or WebM containers and H.264 or AV1 codecs. Keep File type and Codec on auto and Quality (CRF) at its supplied -1 default for the first run. These settings configure output encoding; they do not change the scene requested from HappyHorse.
7 Dirty Words negative prompt
This is a negative embedding trained on the output you get from entering George Carlin's "7 dirty words" in the positive prompt. It was trained mostly on output from various anime models, so it's likely to work...
Kling two-shot storyboard
Generate one Kling video containing two consecutive storyboard shots, with optional first-frame guidance and native audio. For a first run, replace both shot prompts with your subjects, action, setting and camera movement. Set each shot's duration in seconds; the defaults request two seconds followed by three seconds. Each prompt supports up to 512 characters, and each shot duration supports 1–15 seconds. These controls request the segments and their combined duration; the example prompts describe a boxing sequence. Upload a start image and leave Start image enabled for image-to-video generation. An upload is required while that stage is enabled. For text-to-video generation, bypass Start image; no upload is required. Aspect ratio offers landscape 16:9, portrait 9:16 or square 1:1 in text mode and is ignored when a start image is supplied. Kling V3 is selected by default and supports 720p, 1080p and 4k, with optional native audio. You can also choose Kling 3.0 Turbo, but select 720p or 1080p before running it. Turbo always generates native audio and ignores the audio toggle. This tool supplies exactly two storyboard segments; other segment counts and ordinary single-prompt generation are not configured here. Leave Seed at -1 for a fresh run. A fixed seed does not make the provider's output reproducible. Generation uses authenticated Kling API access, sends prompts and any supplied start image to the provider, and incurs provider charges. The result is returned as one video. Advanced Video output controls offer file type and codec selection; leave both at auto for automatic selection, or choose a compatible explicit combination.
Corneo's Marin Kitagawa TI Embedding
This is a Textual Inversion Embedding of Marin Kitagawa from My Dress-Up Darling (sono bisque doll wa koi wo suru).
Audio-guided human video
Generate one human-focused video from source audio and positive and negative prompts, optionally guided by a reference image. The uploaded audio supplies HuMo conditioning and is also included as the soundtrack. For a first run, upload the desired audio excerpt and describe the person, clothing, setting, framing, and action in Positive prompt. Retain or edit Negative prompt to specify unwanted visual features; its supplied default is in Chinese. Upload a reference image to guide appearance and composition, or bypass Reference image to use audio and prompts alone. The reference stage starts enabled, so explicitly bypass it when no image is supplied. Start with 640 × 640 pixels, 97 frames, 25 FPS, and latent batch size 1. The generated frames represent approximately 3.88 seconds of playback. Prepare an audio excerpt appropriate for that span before uploading: this interface provides no verified audio trimming or time-stretching controls. Width and height use multiples of 16. Frame count follows 1 plus a multiple of 4, such as 97, 101, or 105. Frame count controls the number of generated frames; frame rate controls their playback duration. Changing frame rate does not change audio encoding or stretch the soundtrack. Optional 10-bit video can retain smoother gradients, but downstream codec and player support varies; the default is 8-bit. Advanced sampling exposes seed, steps, CFG, sampler, scheduler, denoise fraction, and sampling shift. Seed -1 randomizes each run; a fixed seed reuses sampling noise. The supplied preset uses six steps, CFG 1, uni_pc, simple scheduling, denoise 1, and shift 8. The fixed Lightx2v step-distillation adapter starts at strength 1; its strength is not a dedicated motion control. Keep these settings initially. The model, text encoder, audio encoder, and VAE remain fixed. The result is one video with the uploaded audio, with automatic file type and codec by default. Advanced Video output settings expose container and codec choices. Larger latent batches still enter the same video assembly rather than producing separate clip outputs. There is no separate image or audio output.
SeedVR2 video restore and upscale
Restore and resize an uploaded video with the fixed SeedVR2 3B Int8 model. The default run processes the full clip at 2× its source width and height with one sampling step. No text prompt is required: conditioning comes from the video itself. The full clip or selected segment supplies the output frame rate, audio when present, and bit depth to video assembly; final encoding also depends on the output settings. For a first run, upload a short video and keep the defaults. Enable Trim video to process a segment. Start time and Duration are in seconds; Duration 0 continues to the end. Require full duration raises an error if the requested duration cannot be provided. Segment settings apply only when trimming is enabled. Scale factor changes both dimensions while preserving aspect ratio: 2 doubles width and height, 1 retains the source dimensions, and values below 1 reduce them. A factor of 2 produces approximately four times as many pixels per frame. Resampling method controls interpolation before restoration; the source recommends lanczos for upscaling, area for downscaling, and nearest-exact for pixel art. Color matching optionally brings restored colors closer to the resized source. lab transfers color in CIELAB space, wavelet transfers low-frequency color, adain matches global channel statistics, and none skips color transfer. Enable Temporal chunking to sample automatically sized latent chunks, then merge them before decoding. Chunk overlap is measured in latent frames, rather than seconds or source video frames; overlapping chunks are crossfaded during merging, while 0 uses plain concatenation. It applies only when chunking is enabled. Chunk sizing is automatic; manual chunk size is not available in this Pipe. Chunking addresses sampling memory, while VAE encoding and decoding have their own spatial tile and temporal window controls. Long clips and large output dimensions can still require substantial memory; use a shorter segment when needed. Denoise 1, one sampling step, CFG 1, euler, and simple are the source restoration defaults. Lower Denoise values change the sampling schedule and may reduce changes; test on a short segment. Other sampling settings depart from the source setup and are experimental. Seed -1 selects a new random seed each run; use a fixed seed when comparing settings. Video file type and Codec default to automatic selection. If choosing them explicitly, use a compatible container and codec supported by your intended player. Quality (CRF) -1 keeps automatic selection; lower nonnegative values request higher encoding quality and generally larger files, with the effective range depending on the codec.
Portrait lighting transfer
Generate a relit portrait or subject image with Qwen Image Edit and the Light Migration LoRA. For a first run, upload the required subject image and a lighting reference, keep the default relighting instruction, and leave Lightning mode off. Start with standard steps 20, standard CFG 4, Light Migration strength 1 and denoise 1. Run the Pipe to receive one edited image. Add subject details you want retained to the positive prompt; the negative prompt can remain blank. For text-directed lighting, bypass the Lighting reference stage and replace the default instruction with a direct description of the desired illumination and colors. Leaving the upload empty alone does not select this mode. The subject image remains required. Lightning mode adds the supplied Qwen Image Edit 2509 four-step Lightning LoRA and switches steps and CFG together. Its defaults are four steps and CFG 1. Standard steps and CFG have no effect while Lightning mode is on; Lightning steps, CFG and LoRA strength have no effect while it is off. Light Migration strength applies in both modes. The required model files must be available, including the four-step Lightning LoRA when using that mode. Denoise ranges from 0 to 1 and controls regeneration of the initial subject latent. Lower values retain more source structure but may weaken the lighting change. Light Migration strength controls the lighting LoRA's model influence, rather than an exact percentage of transferred lighting. Use seed -1 for a fresh random seed or a fixed seed to compare edits with otherwise unchanged settings. Advanced controls include separate steps and CFG for each mode, Lightning LoRA strength, sampler, scheduler, guidance normalization and sampling shift. The source sampler and scheduler defaults are Euler and Simple. Steps must be 1–10000 and CFG 0–100, matching the downstream sampler. Guidance normalization and sampling shift apply in both modes; normalization before CFG can amplify guidance. The subject is automatically scaled before generation, so output dimensions may differ from the upload. There is no independent output-size control. This is generative editing: identity, edges and fine details may change. Uploaded transparency is not preserved.
Style Rust Magic
This is a textual inversion that I made inspired by The Last of Us, created with Dream Artist on a single image that I generated. I was trying to do a post-apocalyptic feel. It adds a lot of gadgets, rusted...
Wan 2.2 5B video generation
Generate a silent video with Wan 2.2 5B from a scene and motion description, optionally using an uploaded image to condition the opening. For a first run, leave Source image bypassed and replace the positive prompt with your scene, subject movement, visual style, and camera movement. Keep the supplied negative prompt and default 1280 × 704 dimensions, 121 frames, 24 FPS, and 20 sampling steps. These settings produce approximately five seconds of video. Seed -1 selects fresh random noise each run; use a fixed seed when comparing changes. To animate an image, enable Source image and upload an opening image. Describe the desired movement and continuation in the positive prompt. The upload connects to the opening-image input, rather than supplying a video for editing. Uploaded transparency is not preserved through the connected RGB image path. Width and height are measured in pixels and use multiples of 32. Frame counts use 4n + 1, such as 49, 81, or 121. Increasing dimensions or frame count increases generation work. FPS controls playback speed: with a fixed frame count, increasing FPS shortens the video rather than generating additional frames. Keep batch size at 1 for one sequence; larger batches still feed one video output and do not provide separate video files. The negative prompt discourages unwanted content and artifacts. Advanced Sampling and Noise schedule controls change the denoising process; begin with their supplied defaults and denoise 1. Lower denoise is not a direct opening-image fidelity control. Use 8-bit video for broader playback compatibility. The 10-bit option requires compatible encoding and players. Advanced Video output controls select container, codec, and compression quality. Start with automatic file type and codec and quality -1; explicit encoding choices must be compatible. Encoding quality affects compression, not generated scene content. The output has no audio, even when the prompt describes music or sounds.
SCG Embedding Toolkit for Version 1.5 Models - A collection of tools and helpers from my personal kit
Over the last 6 months, I've created quite the collection of helper and tool embeds for doing things like fixing faces, pushing realism, creating anime scenes, making subjects smile, or cry, or face the camera,...
