SeaArt ComfyUI Update Announcement | 2026-03-12 | Issue 03
Thank you for your continued attention and support!
In this issue, we will share recent ComfyUI node and resource updates with you, so you can stay informed of the latest additions in a timely manner. New in this issue:
Paid Nodes (Points)
Note: The following are paid nodes on the SeaArt platform. Point consumption will vary depending on parameters such as resolution, duration, and mode. For the exact point cost, please refer to the real-time display on the corresponding workflow page.
ViduQ3-Turbo Series
●ViduQ3Turbo_Text2Video(Cost:25-800)
●ViduQ3Turbo_Image2Video(Cost:25-800)
Workflow example entry: https://www.seaart.ai/workFlowDetail/d6ijc95e878c73cgv0ng
SeaArt-Nano Banana 2
Workflow example entry: https://www.seaart.ai/workFlowDetail/d6iilmte878c73atd02g
SeaArt FilmEdit 3.0
Workflow example entry: https://www.seaart.ai/workFlowDetail/d6igd2te878c73at3i40
Kling-V3-Omni Series
●KlingV3Omni_Text2Video(Cost:255-2175)
●KlingV3Omni_Image2Video(Cost:255-2175)
●KlingV3Omni_TI2I(Cost:40-640 Per 1 Batch)
Workflow example entry: https://www.seaart.ai/zhCN/workFlowDetail/d643dele878c73djqq30
Plugin Nodes
1. ComfyUI_SeedVR2_Tiler
Used for tiled upscaling and seamless stitching of large images. It is suitable for processing high-resolution images together with the SeedVR2 model, improving upscaling detail performance while reducing VRAM pressure.
2. Anima Mod Guidance
A guidance enhancement node for Anima models. By introducing additional CLIP modulation signals, it usually helps improve prompt adherence and alleviates some color leakage issues.
3.ComfyUI-Flux2Klein-Enhancer
A prompt enhancement node for FLUX.2 Klein. It allows fine adjustment of text and reference image weights, which usually helps improve prompt following and image editing controllability.
4.ComfyUI-VFI
A video frame interpolation node that can generate intermediate frames on demand within workflows, used to improve video frame rate and motion smoothness while balancing speed and effect adjustment.
5.ComfyUI-TeleStyle
Used for video style transfer. It can stylize video content based on a reference style image and is suitable for creating video content with a unified visual style.
Resource Updates (Models / Static Resources, etc.)
MarigoldModelLoader node
New ControlNet models added
ltx2.3 related resource updates
AudioEncoderLoader now supports Japanese wav2vec2 audio models
The required model resources for Wav2Vec2 Model Loader have been added, resolving the issue where no models were available in the node dropdown list.
The newly added/updated model resources in this issue are as follows: (all can be found on the model page of the SeaArt ComfyUI official account)
FireRed-Image-Edit (Checkpoint)
ace-step-1.5 base(Checkpoint)
Wan21-14B-SCAIL-preview_fp8_scaled_mixed(Checkpoint)
FireRed-Rapid-AIO-SFW(Checkpoint)
LTX-2.3-22b(Checkpoint)
FireRed-Image-Edit-1.1-GGUF(Checkpoint)
sdpose_wholebody(Checkpoint)
FireRed-Image-Edit-1.0-Lightning-8steps（lora）
FireRed-Image-Edit-LoRA-Zoo（lora）
ltx-2.3-22b-distilled-lora-384（lora）
ltx-2.3-22b-distilled-lora-dynamic_fro09_avg_rank_105（lora）
Wan22_I2V_VBVR_HIGH_rank_64（lora）
Wan22_T2V_UnifiedReward-Flex_LoRA_bf16（lora）
More model and workflow collections: https://www.seaart.ai/zhCN/user/M001 (recommended to bookmark/follow; future updates will be compiled there)
