Extensions/ComfyUI-MAINodes
ComfyUI Extension

ComfyUI-MAINodes

MatlowAI's MiniMax-H3 nodes: Dials (training-free steering sliders, rack, time curves), Contact-Sheet diffusion, and Motion Lab (test-time de-roping of fast motion via oracle-guided v2v time-smear)

By matlowai·Created 3 months ago·Updated 4 days ago· 250
matlowai/ComfyUI-MAINodes
Nodes105
On cloudLocal install
Categoryconditioning/minimax/motion, model_patches/minimax/motion
Stars250
Updated4 days ago

Nodes (105)

H3 Add Latent Guide (aligned latent rows) [experimental]
conditioning/minimax/motion
H3 Attention Measure
model_patches/minimax/motion
H3 Audio Recover (hold-map atempo, pitch kept)

The audio half of the de-rope fix, where dialogue goes to survive

audio/minimax/motion
H3 Audio Smear (hold-map stretch, pitch kept) [alpha]

The audio seed that keeps dialogue honest through a de-rope

audio/minimax/motion
H3 Capability Probe (alpha)

Ask the running ComfyUI what it can actually do for H3

MAINodes/alpha
H3 Clock Remap (any model, presets)

De-rope any model, not just MiniMax H3

image/minimax/motion
H3 Conditioning Bank (encode once, no TE per item) [alpha]

Encode your prompt once, stop paying for the text encoder every time

conditioning/minimax
H3 Cond Strength (per kind)
model_patches/minimax
H3 Contact Sheet (five views)

Five views of one subject, packed into one H3 latent

model/conditioning/minimax
H3 Contact Sheet Decode

The mandatory other half of the contact sheet

latent/minimax
H3 Core Compatibility (alpha)
MAINodes/alpha
H3 Delta Color Carry (alpha)

Stop the chain of clips from slowly getting darker

MAINodes/alpha
H3 De-rope Sigmas (alpha)
MAINodes/Dials
H3 Drawn Plan (alpha)

Draw the timing envelope in the web editor, wire it into the graph

latent/minimax/timeline
H3 Drift Control (alpha)

The node that stops chained segments from going mushy

MAINodes/alpha
H3 DyRoPE (layer-wise / sigma-faded time geometry) [experimental]

Give the true clock only to the blocks that can afford it

sampling/custom_sampling/samplers
H3 Evict Text Encoder (unload after encode)

Shove the 15 GB text encoder out of VRAM the moment it's done

MAINodes/VRAM Lab
H3 Exact Recover (24fps frame selection)

The boring node that makes the de-rope honest

image/minimax/motion
H3 Expert Schedule (base head, turbo tail)

Base model starts the structure, turbo finishes the polish

sampling/custom_sampling/schedulers
H3 Extension Plan (alpha)

Do the grid math for chaining H3 clips so you don't have to

MAINodes/alpha
H3 Fake Quant (A6: simulate NVFP4/FP8 activations per region, alpha)

Pretend the activations are 4-bit, measure what breaks

MAINodes/VRAM Lab
H3 Free Cache (empty allocator between stages)

A 17 GB refund you can collect with one node

MAINodes/VRAM Lab
H3 Guide Anchor (pin / float)
conditioning/minimax
H3 Hold Map Remap (scale / floor / ceiling) [alpha]
latent/minimax/motion
H3 Indecision Oracle (x0-jitter / blend) [experimental]

A second oracle that reads the model's doubt, not the motion

latent/minimax/motion
H3 Inject Schedule (v2v sigmas, 0.70)

The one knob that decides if your de-rope actually fixes the smear

sampling/custom_sampling/schedulers
H3 Jerk Heatmap (oracle overlay tile)

Watch the de-rope oracle think, then trust it a little less

image/minimax/motion
H3 Jerk Oracle (profile / window / hold map)

The oracle that reads H3's latent and finds the frames that smear

latent/minimax/motion
H3 Latent Bank (sample once, skip the pass) [alpha]

Keep the pass-1 latent so window items stop re-rendering it

latent/minimax
H3 Latent Upscale (video only, audio kept) [experimental]

Upscale the video latent, keep the audio bit-exact, skip the VAE round trip

latent/minimax/motion
H3 Load Hold Map (sidecar)

Steal a hold map from another model's graph without loading that model

image/minimax/motion
H3 Manual Hold Map (ranges to holds, gate) [alpha]

Tired of the oracle being overzealous? Type the ranges yourself

latent/minimax/motion
H3 Mask Conversion (#15988)
model_patches/minimax/motion
H3 Memory Probe (ledger + allocator trace, alpha)

Find out what's eating VRAM, block by block, without guessing

MAINodes/VRAM Lab
H3 Mid Insert (change the token grid MID-denoise) [experimental]

Insert token-times mid-denoise, so the coarse pass and the dense pass share one trajectory

latent/minimax/motion
H3 Mode Switch (preview / final, lazy)

One graph, two pipelines, a dropdown that only runs the branch you picked

utils/minimax
H3 Motion Composite (subject regen, background baseline)

Keep the subject regenerated, hand the background back to the baseline

image/minimax/motion
H3 Motion Editor (timeline, masks, automation) [alpha]

Paint where the de-rope should act

latent/minimax/motion
H3 Plan Estimate (alpha)

Price a plan before you run it — from your own recorded renders, not a spec sheet

latent/minimax/timeline
H3 Plan Settings (alpha)

Pull a plan's execution settings out as typed wires, so the graph can't drift

latent/minimax/timeline
H3 Precision Probe (A6: 4-bit vs 8-bit activation map, alpha)

Which activations could survive 4-bit? This node measures before you commit

MAINodes/VRAM Lab
H3 Prefix Freeze Mask (alpha)

Give the carried prefix a freeze mask so pass 2 doesn't re-texture the seam

MAINodes/alpha
H3 Probe Schedule (early-oracle head)

Read the oracle early and don't pay for the rest

sampling/custom_sampling/schedulers
H3 Protect Prefix (alpha)

Stop the de-rope from retiming the continuity anchor you're carrying over

MAINodes/alpha
H3 Flight Recorder Start (alpha)

The front half of the flight recorder that calibrates your cost estimates

latent/minimax/timeline
H3 Flight Recorder Stop (alpha)

Read the peaks, append the line, pass the wire on

latent/minimax/timeline
H3 References (alpha)
MAINodes/H3
H3 Reference to Video (refs) (alpha)
MAINodes/H3
H3 Repair Plan (alpha)

'Frames 45 to 47 are wrong' → the whole arithmetic of a repair pass

MAINodes/alpha
H3 Repair Splice (alpha)

Stitch the repaired frames back in, and prove nothing else changed

MAINodes/alpha
H3 Retime Audio (bullet-time bed + transients) [alpha]
audio/minimax/motion
H3 Save Hold Map (sidecar)

A de-rope render without its hold map is a secret that dies when you close the graph

image/minimax/motion
H3 Scene Color Stats (alpha)

The thermometer for colour drift across chained segments

MAINodes/alpha
H3 Seam Normalize (alpha)

The hidden prefix is a calibration chart, and this node reads it

MAINodes/alpha
H3 Segment Crop (regen only the window) [alpha]

Stop regenerating the frames that were already fine

image/minimax/motion
H3 Segment Splice (crossfade reassembly) [alpha]

Put the clip back together so nobody can find the cut

image/minimax/motion
H3 Streamed Blocks (exact low-VRAM, alpha)

The de-rope pass on a 16 GB card, bit-equal to the 96 GB one

MAINodes/VRAM Lab
H3 Tail Context (alpha)

How the next segment gets to know how this one ended

MAINodes/alpha
H3 Temporal Insert (insert token-times, freeze originals) [experimental]

The de-rope done in latent space, so your good frames never get re-rendered

latent/minimax/motion
H3 Timeline Analyze (alpha)

Turn two oracles' opinions into a retiming plan the graph obeys

latent/minimax/timeline
H3 Timeline Render (alpha)

Compile a plan into a legal graph, and price it before you queue

latent/minimax/timeline
H3 Time Smear (integer holds)

The node that teaches MiniMax-H3 to draw fast motion twice as slowly

image/minimax/motion
H3 Trajectory Bank (checkpoint every step)

Checkpoint your sampling run every step, so branching isn't a re-run

sampling/custom_sampling/samplers
H3 Trajectory Load (branch from a step)

Resume a sampling run at step 18, not step 0

latent/minimax/motion
H3 Trim (alpha)

Throw away the hidden prefix without throwing away your timeline

MAINodes/alpha
H3 True Clock (density-corrected RoPE t-grid) [experimental]

Why your held frames play back at double speed, and the fix that's off-distribution

sampling/custom_sampling/samplers
H3 V2V Init (nested AV latent)

The node that turns a slowed-down clip into a regen ready to go

latent/minimax/motion
H3 Video Fit (source clip -> 17k+5 frames) [alpha]

Why your MP4 doesn't run through H3, and the node that fixes it

image/minimax/motion
H3 Window Collect (bank windows, splice when full) [alpha]

Bank every window to disk so an OOM costs you one frame, not the run

image/minimax/motion
H3 Window Crop (internal)

The internal node you'll touch only when a window plan is lying to you

image/minimax/motion
H3 Window Expand (rolling regen)

Rolling regeneration for clips too long for your VRAM

image/minimax/motion
H3 Window Loop (rolling regen, one node) [alpha]

The whole de-rope chain, inside one node, on a VRAM leash

image/minimax/motion
H3 Window Pass (internal)

The internal no-op you'll see in a windowed graph

image/minimax/motion
H3 Window Plan (split the pass into N windows) [alpha]

Cut a de-rope pass into N affordable windows

image/minimax/motion
H3 Window Seam (internal)

Pin-and-trim where two windows meet mid-burst

image/minimax/motion
MAI Concept Capture Arm (alpha)

Tap a model's insides and see what a reference actually does

MAI/concept (alpha)
MAI Concept Capture Flush (alpha)

The 'now file it' node of a Concept Lab capture

MAI/concept (alpha)
MAI Concept Inject Delta (alpha)

Put a measured capture back into the model

MAI/concept (alpha)
Combine Dials (alpha)
MAINodes/Dials
Dial Hold Map (alpha)
MAINodes/Dials
Make Dial (alpha)
MAINodes/Dials
Make Dial from Pictures (alpha)
MAINodes/Dials
Make Dial from Videos (alpha)
MAINodes/Dials
Dial Rack (alpha)
MAINodes/Dials
Condition
MAINodes/Flow
Filter (list)
MAINodes/Flow
Gate (process if)
MAINodes/Flow
LLM Choose
MAINodes/AI Decisions
LLM Judge
MAINodes/AI Decisions
Partition (list)
MAINodes/Flow
Flow Probe
MAINodes/Flow
Safe Function
MAINodes/Flow
Lazy Select
MAINodes/Flow
MAI Load Video (path)

A video in, by path, as ComfyUI's VIDEO type

image/minimax/video
Render Preset (alpha)
MAINodes/Dials
MAI Seed Hunter (compare + winner seed)

Find the keepable seed by watching 2-6 candidates side by side

image/minimax/video
MAI Select Every Nth

The uniform frame decimation for the no-hold-map case

image/minimax/video
MAI Video Compare (2-6, synchronized)

2-6 renders, synchronized, in the browser, for zero VRAM

image/minimax/video
MAI Video Out (file + sidecars + draft preview)

Save a render with its provenance sitting right beside it

image/minimax/video
RelayVSR • Block Loop Lab (Experimental)
RelayVSR/Research
RelayVSR • X2 / X4 Settings
RelayVSR
RelayVSR • Upscale IMAGE Sequence
RelayVSR
RelayVSR • Upscale Video File
RelayVSR
RelayVSR • Upscale VIDEO
RelayVSR
RelayVSR • Decode Video to IMAGE Batch
RelayVSR/Utilities
Readme

ComfyUI-MAINodes

Custom nodes for MiniMax H3 video. Three parts:

  • Motion Lab fixes fast-motion smear (backflips, sword arcs, whip-fast turns) at inference time by re-rendering the clip as a slowed-down copy of itself and dropping the held frames afterward. No extra model, no training.
  • Dials (alpha) are training-free steering sliders: make one from two prompts or two groups of pictures, stack them in a Dial Rack, and change them over the clip. Also runs on MOSS-Image, and with limits on Krea 2.
  • Contact-Sheet diffusion makes five views of one subject from one reference image.

Around them: a video compare player, low-VRAM nodes for 16 to 24 GB cards, and research nodes marked (alpha).

baseline vs inject 0.70 vs inject 0.50

The baseline (left) smears the aerial spin; the two de-roped versions keep it clean with the same choreography. Default settings, no per-clip tuning. More demos: Motion Lab page, Dials page.

Like it? A star helps. Want to feed the GPU? Sponsor the experiments.

Start here

| you want to | start from | read | |---|---|---| | make an H3 video (recommended start): Pika H3 Flex + de-rope; steer look, mood or motion with Dials sliders | examples/dials/dials_simple_pika_derope.json | dials/README.md | | fix smeared fast motion (Motion Lab graphs) | examples/motion_pipeline_ref2va_audioinit.json | Motion Lab, TUNING.md | | five views of one subject | examples/contact_sheet.json | below | | upscale a video 2x or 4x (RelayVSR) | examples/relayvsr/workflow_comfyui_native_video.json | relayvsr/README.md | | run on a 16 to 24 GB card | examples/motion_pipeline_lowvram.json | LOWVRAM.md, HARDWARE.md | | compare renders side by side | the MAI Video Compare node | VIDEO_TOOLS.md | | pick among all the graphs | | examples/README.md |

Every graph comes as name.json (drag onto the canvas) and usually name_api.json (the same graph for scripts). Most Motion Lab graphs need ComfyUI-KJNodes.

Motion Lab (de-rope)

One latent token spans four video frames, and at high speed those four frames need four poses one token cannot hold, so H3 smears. The pipeline reads where motion runs too hot from the clip's own latent (the oracle), holds those frames so the model has room, regenerates video-to-video from that slowed init at partial denoise, and recovers the original frame rate by exact frame selection.

(baseline video) -> VAEDecode frames        (baseline latent)
        |                                        |
        v                                        v
   H3TimeSmear  <-- hold_map ------------- H3JerkOracle
        |  (integer holds)                       |
        v                                        |
    VAEEncode -> H3V2VInit -> SamplerCustomAdvanced
                                  ^
              H3InjectSchedule ---/
                                  |
                              VAEDecode -> H3ExactRecover -> original fps

Start from motion_pipeline_ref2va_audioinit.json: pass 1 at 12 steps on the base model, the turbo LoRA at 6 steps for the de-rope, dialogue kept intact by the audio init. 192 frames at 1 MP take about 12 minutes end to end on an RTX PRO 6000 (36 for 25 steps both passes without turbo). Reach for inject (0.50 sharper, 0.70 safer) before any other setting, and try a 2 to 3 second clip first.

The full guide, with every graph, node table, timing table and the alpha features (repair, window mode, DyRoPE, chaining, the editor), is docs/MOTION_LAB.md.

Dials (alpha)

Training-free steering sliders for MiniMax H3 (also MOSS-Image, and Krea 2 with limits). A dial is a ~14 KB file made from the difference between two prompts (or two groups of pictures); a negative strength pushes every generation toward the first side, a positive one toward the second. No training, about a minute to make one. Stack them in a Dial Rack, see how they interact, and vary them over the clip. Full guide: dials/README.md; what we learned: dials/LESSONS.md; what's next: dials/ROADMAP.md; dial library on Hugging Face. See it in action: the Dials project page and demo video. The original stable release, with the simple Load/Apply Dial nodes, is ComfyUI-Dials; both packs share the same dial files.

What you can do with them

| capability | what it gives you | |---|---| | Make a dial from prompts | A ||| B pairs, prompt sets with weights, hold pairs (!) that keep a quality fixed | | Make a dial from pictures or clips | two groups of pictures of any size (or matched video clips); pick how they are read: content, look, bare, all averaged or your own sentence | | Dial Rack | any number of dials on one model: on/off, strength, solo, and H to hold a direction out of the mix. Row order should not change the result (one render did; under investigation) | | See what dials do | a 2D map of two dials with their useful and overcooked ranges, an interaction matrix (who reinforces or cancels whom), a "who pushes" bar, and an influence wheel over 14 probe axes (motion, light, colour, mood, voice...) | | Edit a dial | open a dial, see its probe profile, drag bars to add or remove a component (say, take the colour out of a dial), and export a new dial with its parentage recorded; Combine Dials mixes two dials into one file | | Control it over time | draw a strength curve over the clip with the mouse (click to add a point, drag, double-click to delete, snap to seconds or beats) or drive it from audio; the dial switches inside one generation, not in a second pass | | Control it over depth | a 50-cell strip sets how much of the dial each DiT block gets | | Keep audio clean | an "audio: protect" option leaves the soundtrack undialled (measured: speech errors 0.29 to 0.00 in 3 of 4 cases) | | Works with de-rope | a hold-map link keeps time curves in real clip time through a de-rope run | | Other models | the same nodes detect MOSS-Image (works, about 2-4x H3's strengths) and Krea 2 (partial); dial files remember their model family | | Examples and help | a Simple front door (Render Preset, Dial Rack, one blueprint; core-only and KJNodes variants) and Pika H3 Flex + de-rope workflows for the rack, picture dials (folders, two pictures, groups) and prompt sets in examples/dials/; Node Info guides on the Dial Rack, Make Dial from Pictures, Render Preset and H3 De-rope Sigmas |

Honest limits. Dials are alpha. Strengths are starting points measured on Pika H3 Flex's 6-step schedule, mostly one seed. Time curves switch smoothly (a lighting dial ramps over about 1.5 s; colour and style dials switch much faster) and cannot follow individual drum hits. Picture dials come out weaker than text dials. Krea 2 steers on one side only. A dial can leak into other qualities (a "psychedelic" dial also brings rave outfits); the lessons file lists the known ones.

Quick start. Load examples/dials/dials_simple_pika_derope.json (core ComfyUI only; _kj adds KJNodes' VRAM savers and folder loader): pick a Render Preset (Pika, Pika + reference pictures, DMAD, or your own), write a prompt, drag a slider on the Dial Rack, and queue; de-rope is one choice on the same node (off, oracle or flat; 79 s off at 1280x704). dials_rack_pika_derope.json is the same render with every node in view. To make your own, add Make Dial, write two prompts that differ in one quality, and wire its output into a rack row.

Contact-Sheet diffusion

Five standalone image latents packed on the model's time axis, jointly denoised, decoded independently. Use with a Turnaround LoRA from matlod/minimax-h3-turnaround. Nodes: H3 Contact Sheet, H3 Contact Sheet Decode. Drag-in workflow: examples/contact_sheet.json (API twin alongside) — point the LoadImage at your reference image, pick your downloaded LoRA file, queue. ref_image is optional: leave it unconnected for a text-only sheet (drop the <Picture 1> tag from the prompt); examples/contact_sheet_t2i.json is that graph. Stock loaders and sampler throughout; 28 steps of res_multistep at denoise 1.0, LoRA strength 0.75. A scripted example is in example_api_workflow.py. Previously published as ComfyUI-H3-ContactSheet; that repo remains up for existing installs.

More

  • Video tools: MAI Video Out (file plus hold-map and metadata sidecars), MAI Video Compare (2 to 6 renders, synchronized, with frame-exact export), MAI Seed Hunter, and a standalone compare deck. VIDEO_TOOLS.md.
  • Small cards: H3 Streamed Blocks, H3 Memory Probe, H3 Free Cache and friends render a long de-rope that OOMed a 96 GB card on a 16 GB card in a 32 GB machine. LOWVRAM.md, speed dials in FASTER.md.
  • De-rope for any model (alpha): H3 Clock Remap retimes one hold map onto another model's clock from a preset (minimax-h3, ltx-2.5, Wan 2.2). The LTX-2.5 de-rope deck behind https://matlowai.github.io/flipbook/derope.html was built with it. DEROPE_ANY_MODEL.md.
  • What is stable: the Motion Lab pipeline and the contact sheet. Concept Lab, the timeline surface, the motion adapter, manual masks, the audio init and Dials are alpha: ALPHA.md says what each can do and where it was tested. Alpha nodes carry (alpha) in the menu and load behind a guarded loader, so a broken one cannot take the pack down.
  • Snap (pre-alpha): a snap-together view over the ComfyUI graph, with agent tools. It works but is not polished yet. How to try it, known issues and the third-party model licenses: docs/snap/README.md. It is off until you turn it on (Extensions menu: "Snap (pre-alpha): turn on"), and off it adds nothing to ComfyUI but that one entry.
  • Node Shelf: hide node packs, or single nodes, you don't use. Open it from ComfyUI's command palette ("Node Shelf"), the Extensions menu, or Snap's top bar. Hidden nodes leave node search, the node library, the right-click Add Node menu and Snap's Add brick; they stay installed, so workflows that use them still load and run. Needs a ComfyUI interface from about August 2025 or later; on an older one the dialog says so and hides nothing. Also available by itself: comfyui-node-shelf (installing both is fine: one copy runs).
  • What's new: RelayVSR upscaling with quality presets and 16 / 24 GB tiers (relayvsr/README.md, 2026-11-03). Older notes in docs/NEWS.md; open questions in ROADMAP.md.

Install

cd ComfyUI/custom_nodes
git clone https://github.com/matlowai/ComfyUI-MAINodes

Most Motion Lab example graphs also need ComfyUI-KJNodes (the Dials picture-folder example uses its LoadImagesFromFolderKJ too). They use PathchSageAttentionKJ, MiniMaxH3MemoryEfficientSageAttentionPatch and MiniMaxChunkFeedForward from it. Without KJNodes a graph loads with those three missing and will not run. The nodes in this pack themselves have no such dependency.

Cloned before 2026-08-19 and git pull errors? Run git fetch origin && git reset --hard origin/main, or re-clone.

Restart ComfyUI. Nodes appear under latent/minimax/motion, image/minimax/motion, sampling/custom_sampling/schedulers and MAINodes/Dials.

Support MAINodes

MAINodes is free and open source. If it saved you time, fixed a workflow or made something cool possible, you can help fund the next round of GPU time, benchmarking and increasingly questionable experiments:

License

GPL-3.0-or-later. Copyright (C) 2026 MATLOWAI. See LICENSE.