ComfyUI ecosystem updates — open-source model releases, custom nodes, workflows, and tools for image, video, and audio generation.
Text to speechYuE2-3B ComfyUI workflow: the MAP team's open music generation model turns lyrics into songs with editable scores, beating Suno v5 on WildSongBench. Weights now in master.
CC-BY-NC-4.0
drbaph extracted the VDN-H3 Turbo adapter into a standalone 8-step LoRA for MiniMax H3, with pruned fl2va/ref2va variants plus an INT8 ConvRot VDN checkpoint for Comfy Kitchen.
minimax-h3-community-license-agreement
EcosystemWAS Node Suite v3 rebuilds the classic ComfyUI node pack: 463 nodes, no automatic pip installs, torch-native processing, Three.js 3D scenes and in-node video compare.
Text to imagecomfyui-SelfLift runs early denoising at low resolution and finishes at full res, a training-free speedup for Z-Image, FLUX.2-Klein and MiniMax H3 with TST correction.
Text to videoAlibaba's TaoLive AIGC team open-sources TaoMate-H3, a 3-step LoRA runtime that streams synchronized audio-video in chunks for 35 FPS long-form MiniMax H3 generation.
minimax-h3-community-license-agreement
Text to videoAI Movie Studio 2 is an open-source AI filmmaking workstation that drives local ComfyUI with a 3D storyboard, continuity system, Long Take mode and timeline export.
Text to speechRumik AI open-sources rumik-oss 1, a 3B text-to-speech model for 22 Indic languages plus English, with code-switched synthesis, delivery conditioning, and inline laughter tags.
cc-by-nc-4.0
Video editingAlissonerdx's H3 sharpness LoRA restores crisp detail in a same-resolution v2v pass: guide-latent conditioning, one 1.2 GB file, native ComfyUI workflow included.
Apache-2.0
Text to videoKijai's block sparse attention lands in ComfyUI core with three backends for MiniMax H3, plus a new memory compiler that cuts VRAM waste and speeds up long H3 runs.
Image to videoA ComfyUI node transplants the FastH3 VSA gate onto MiniMax H3 Ref2VA, keeping reference tokens dense while pruning video tiles for fast 4-step character video.
Download WarmBloodAban's Minimax-h3_Singularity_ref2va_v1.3_int8.safetensors plus the Pruned build: a MiniMax H3 HDR fusion fine-tune for ComfyUI action and VFX video.
apache-2.0
Image to imageA new custom node turns ComfyUI into a layered paint program: selections, SAM-assisted picking, inpaint round trips, filter layers and PSD export in one editor.
Image to videoRefMods are 1 MB conditioning adapters for MiniMax H3: encode a character once, load it like a LoRA in ComfyUI, and reuse it across unlimited generations.
Apache-2.0
Text to videoAn independent distillation project transfers SenseNova U1.5 semantics into MiniMax H3's conditioning space: an 11MB standalone adapter with ComfyUI nodes and workflows.
minimax-h3-community-license-agreement
Text to imageInclusionAI releases LLaDA-Image, a 6B unified generation and editing model with a 4-step Turbo variant, plus a community ComfyUI node pack with INT8 and GGUF weights.
Text to videoDownload minimax_h3_fl2v_turbo_4step_v1.2_768p_comfyui_bf16.safetensors: Lightx2v's 4-step MiniMax H3 Turbo update with cleaner, more stable audio generation.
Text to imageComfyUI adds two official Meta Muse Image API nodes: text-to-image and image edit with up to 10 input images, 14 aspect ratios, and a reasoning strength control.
Text to imageComfyUI's Recraft nodes add V4 Styles and Styles Pro models to the Create Style node, pairing custom trained styles with a style reference workflow for consistent branding.
Image to videoDownload minimax_h3_ref2v_turbo_8step_v1.0_comfyui_bf16.safetensors: Lightx2v's 8-step 768p Ref2VA distillation for MiniMax H3 with audio, plus updated example workflows.
Text to imagecapitan01R's Krea2T Enhancer adds an Attention-Weighted Phrases node for Krea 2: boost or suppress individual prompt phrases directly in the DiT attention.