14Graph nodes
13Node classes
16Inputs
6Models
Tags: videomeditationanimationrelaxationambientcalm
Inputs (16)#
The typed parameter surface callers bind when they request this workflow. Enum options and numeric bounds are the values the workflow document declares.
scene_typeenumdefault zen_gardenType of meditation scene
options:
zen_gardenforest_streamocean_wavesmountain_vistalotus_pondstarry_nightsacred_templecrystal_cavefloating_cloudsaurora_skytime_of_dayenumdefault golden_hourTime of day lighting
options:
dawnmorninggolden_hourdusknighttimelessweather_moodenumdefault clearWeather and atmospheric mood
options:
clearsoft_mistgentle_rainlight_snowethereal_glowmotion_styleenumdefault subtle_movementCamera and element motion style
options:
slow_pangentle_driftsubtle_movementbreathing_rhythmflowing_elementsstatic_with_particlesmotion_intensityenumdefault gentleIntensity of motion
options:
minimalgentlemoderateinclude_particlesbooleandefault trueInclude floating particles (light motes, leaves, etc.)
color_paletteenumdefault naturalColor palette for the scene
options:
naturalwarm_goldencool_sereneearth_tonescosmicchakra_alignedduration_secondsintegerdefault 8min 2max 20Video duration in seconds (frame count = duration * fps; capped for AnimateDiff sliding-context sampling)
fpsintegerdefault 8max 24Frames per second (AnimateDiff motion is trained near 8fps; use interpolation downstream for higher rates)
resolutionenumdefault 720pOutput resolution
options:
480p720p1080ploop_seamlessbooleandefault trueCreate seamless looping video
seedintegerdefault -1Random seed
frame_upscalebooleandefault falsePer-frame 2x enhancement before encoding (RealESRGAN 4x then 0.5 resample); compute-heavy, off by default
checkpointenumdefault dreamshaper_8SD1.5 checkpoint: DreamShaper 8 finetune (default, baked VAE) or base 1.5
options:
dreamshaper_8sd15_basemotion_adapterbooleandefault trueAnimateDiff v3 domain-adapter LoRA (cleans motion-training artifacts)
smooth_motionbooleandefault falseRIFE 4.9 frame interpolation (2x): doubles effective fps before encoding
ComfyUI node graph (14)#
The executable ComfyUI prompt graph: 14 nodes across 13 distinct node classes, wired by 16 data dependencies. Nodes tinted green come from a custom node pack this workflow declares; the rest are ComfyUI core / baked-community classes.
flowchart TD
n0["1: CheckpointLoaderSimple"]
n1["2: CLIPTextEncode"]
n2["3: CLIPTextEncode"]
n3["4: EmptyLatentImage"]
n4["5: ADE_StandardUniformContextOptions"]
n5["6: ADE_AnimateDiffLoaderGen1"]
n6["7: KSampler"]
n7["8: VAEDecode"]
n8["9: VHS_VideoCombine"]
n9["fups_loader: UpscaleModelLoader"]
n10["fups_apply: ImageUpscaleWithModel"]
n11["fups_scale: ImageScaleBy"]
n12["ad_adapter: LoraLoaderModelOnly"]
n13["rife: RIFE VFI"]
n0 -->|clip| n1
n0 -->|clip| n2
n12 -->|model| n5
n4 -->|context_options| n5
n5 -->|model| n6
n1 -->|positive| n6
n2 -->|negative| n6
n3 -->|latent_image| n6
n6 -->|samples| n7
n0 -->|vae| n7
n11 -->|images| n8
n9 -->|upscale_model| n10
n13 -->|image| n10
n10 -->|image| n11
n0 -->|model| n12
n7 -->|frames| n13
classDef custom fill:#12b886,stroke:#0b6b4f,color:#04110c;
class n4,n5,n8,n13 custom;
Nodes (14)#
1CheckpointLoaderSimplecoreckpt_name = {{checkpoint_files[checkpoint]}} tmploutputs:
MODELCLIPVAE2CLIPTextEncodecoretext = {{constructed_prompt}} tmplclip = ◂ node 1 · out[1]outputs:
CONDITIONING3CLIPTextEncodecoretext = ugly, low quality, blurry, text watermark, people, faces, disturbing, scary, rapid motion, chaotic, noisyclip = ◂ node 1 · out[1]outputs:
CONDITIONING4EmptyLatentImagecorewidth = {{resolution_map[resolution].width}} tmplheight = {{resolution_map[resolution].height}} tmplbatch_size = {{duration_seconds * fps}} tmploutputs:
LATENT5ADE_StandardUniformContextOptionscustom packcontext_length = 16context_stride = 1context_overlap = 4fuse_method = pyramiduse_on_equal_length = falsestart_percent = 0.0guarantee_steps = 1outputs:
CONTEXT_OPTIONS6ADE_AnimateDiffLoaderGen1custom packmodel = ◂ node ad_adapter · out[0]model_name = v3_sd15_mm.ckptbeta_schedule = autoselectcontext_options = ◂ node 5 · out[0]outputs:
MODEL7KSamplercoremodel = ◂ node 6 · out[0]positive = ◂ node 2 · out[0]negative = ◂ node 3 · out[0]latent_image = ◂ node 4 · out[0]seed = {{seed}} tmplsteps = 25cfg = 7.0sampler_name = eulerscheduler = normaldenoise = 1.0outputs:
LATENT8VAEDecodecoresamples = ◂ node 7 · out[0]vae = ◂ node 1 · out[2]outputs:
IMAGE9VHS_VideoCombinecustom packimages = ◂ node fups_scale · out[0]frame_rate = {{int(fps) * 2 if smooth_motion else int(fps)}} tmplloop_count = 0filename_prefix = meditation_sceneformat = video/h264-mp4pingpong = {{loop_seamless}} tmplsave_output = trueoutputs:
VHS_FILENAMESfups_loaderUpscaleModelLoadercoremodel_name = RealESRGAN_x4plus.pthoutputs:
UPSCALE_MODELfups_applyImageUpscaleWithModelcoreupscale_model = ◂ node fups_loader · out[0]image = ◂ node rife · out[0]outputs:
IMAGEfups_scaleImageScaleBycoreimage = ◂ node fups_apply · out[0]upscale_method = lanczosscale_by = 0.5outputs:
IMAGEad_adapterLoraLoaderModelOnlycoremodel = ◂ node 1 · out[0]lora_name = v3_sd15_adapter.ckptstrength_model = 0.7outputs:
MODELrifeRIFE VFIcustom packckpt_name = rife49.pthframes = ◂ node 8 · out[0]clear_cache_after_n_frames = 10multiplier = 2fast_mode = trueensemble = truescale_factor = 1.0dtype = float32torch_compile = falsebatch_size = 1outputs:
IMAGEPrompt construction#
templatemeditation scene video, {scene_prompt}, {time_prompt}, {weather_prompt}, {color_prompt}, cinematic quality, smooth {motion_prompt}, {motion_intensity} motion intensity{particles_clause}{loop_clause}, peaceful calming atmosphere, 4k quality, professional cinematography, meditative ambient videoVariables (8)#
scene_prompt{{scene_prompts[scene_type]}} tmpltime_prompt{{time_prompts[time_of_day]}} tmplweather_prompt{{weather_prompts[weather_mood]}} tmplmotion_prompt{{motion_prompts[motion_style]}} tmplmotion_intensity{{motion_intensity}} tmplcolor_prompt{{color_prompts[color_palette]}} tmplparticles_clause{{', floating particles, light motes, drifting elements' if include_particles else ''}} tmplloop_clause{{', seamless loop, perfect cycle' if loop_seamless else ''}} tmplParameter banks (6)#
The prompt / configuration lookup tables this workflow keys into from its inputs — the vocabulary that turns a style / palette / preset selection into graph parameters.
scene_prompts (10)#
zen_gardenpeaceful Japanese zen garden, raked sand patterns, carefully placed stones, bamboo, moss, minimalist beauty, contemplative spaceforest_streamserene forest stream, gentle flowing water, moss-covered rocks, dappled sunlight, ferns, woodland tranquilityocean_wavescalm ocean waves, gentle surf, sandy beach, horizon line, peaceful seascape, rhythmic watermountain_vistamajestic mountain vista, distant peaks, alpine meadow, vast sky, panoramic serenity, elevated perspectivelotus_pondtranquil lotus pond, floating lotus flowers, lily pads, still water, reflections, sacred bloomstarry_nightstarry night sky, Milky Way visible, celestial wonder, cosmic peace, night meditation, stargazingsacred_templeancient sacred temple interior, candlelight, incense smoke, carved stone, spiritual architecture, holy spacecrystal_caveluminous crystal cave, natural gems, underground wonder, light refracting, geological beauty, earth's treasurefloating_cloudsabove the clouds, cloud sea, floating perspective, sky realm, heavenly view, endless whiteaurora_skyaurora borealis sky, northern lights, dancing colors, arctic wonder, cosmic light show, ethereal displaytime_prompts (6)#
dawnearly dawn lighting, soft pink and gold, first light, awakening energy, gentle sunrisemorningfresh morning light, bright but soft, clear sky, new day energy, crisp illuminationgolden_hourgolden hour warm light, magical glow, long shadows, amber tones, photographer's lightduskdusk twilight lighting, purple and orange, setting sun, transitional light, evening calmnightnighttime lighting, moonlight, stars visible, nocturnal peace, quiet darknesstimelesstimeless ethereal lighting, no specific time, otherworldly glow, spiritual lightweather_prompts (5)#
clearclear weather, no obstruction, pure atmosphere, clean air, perfect visibilitysoft_mistsoft mist atmosphere, gentle fog, dreamy haze, mysterious depth, veiled beautygentle_raingentle rain falling, soft precipitation, rain drops, wet surfaces, cleansing waterlight_snowlight snowfall, snowflakes drifting, winter quiet, white dusting, cold serenityethereal_glowethereal glowing atmosphere, magical light, supernatural ambiance, spiritual illuminationmotion_prompts (6)#
slow_panslow camera pan, gradual movement, cinematic sweep, smooth travel, revealing scenegentle_driftgentle camera drift, subtle float, dreamy movement, organic sway, natural flowsubtle_movementsubtle element movement, grass swaying, water rippling, leaves trembling, alive stillnessbreathing_rhythmbreathing rhythm motion, expand and contract, meditative pulse, living scene, yoga breathflowing_elementsflowing element motion, water moving, clouds drifting, wind visible, natural flowstatic_with_particlesmostly static scene, floating particles only, dust motes, light specs, minimal motioncolor_prompts (6)#
naturalnatural color palette, realistic tones, true to life, organic colorswarm_goldenwarm golden palette, amber yellow, sunset tones, cozy warmth, invitingcool_serenecool serene palette, blue and teal, calm tones, refreshing, peaceful coolearth_tonesearth tone palette, brown green ochre, grounded colors, natural materialscosmiccosmic palette, deep purple blue, stellar colors, space tones, infinitechakra_alignedchakra color palette, rainbow spectrum, energy centers, spiritual colorsresolution_map (3)#
480p{"width": 512, "height": 288}720p{"width": 768, "height": 432}1080p{"width": 960, "height": 544}Models & dependencies#
Custom node packs (4)#
The non-core ComfyUI node classes this graph requires; the RunPod worker image the workflow runs on must bake or install a pack that provides every one of them.
ADE_AnimateDiffLoaderGen1ADE_StandardUniformContextOptionsRIFE VFIVHS_VideoCombineCheckpoint aliases (2)#
dreamshaper_8DreamShaper_8_pruned.safetensorssd15_basev1-5-pruned-emaonly.safetensorsModels required (6)#
DreamShaper_8_pruned.safetensorsRealESRGAN_x4plus.pthrife49.pthv1-5-pruned-emaonly.safetensorsv3_sd15_adapter.ckptv3_sd15_mm.ckptOutput contract#
What a successful run of this workflow returns.
typevideoformatmp4codech264audiofalseTaxonomy & routing#
How the control plane classifies this workflow — from the committed workflow-taxonomy-registry.json. It drives the consistency / control surface the agentic director can exercise over the workflow.
assetFamilysocial-shortform-vertical-video-packoutputPackageProfilevideo-master-profilecontrolModalitiescamera-lockcamera-path-guidecolor-script-lockdepth-constraintidentity-adapterkeyframe-anchor-locklatent-reusemodel-lockmotion-vector-lockpalette-lockpose-constraintprompt-template-lockreference-ensemblesampler-scheduler-lockscene-lockseed-lock +4 moreconsistencyDimensionsmotionenvironmentlightingcolor-scriptnotesAuto-mapped from video defaults