Workflows are able to Text-to-Image, Multi-Reference Image-to-Image and Inpainting/Outpainting.<h3 id="update-1:">Update 1:</h3><h3 id="comfyui-dlss5-enhancer-will-be-included-in-the-next-v8.0-update!-it-is-a-game-changer-for-image-quality.-visit-their-github-page-to-get-a-jump-ahead-if-you-put-this-workflow-to-good-use.-for-current-workflows-plug-it-into-save-image:-out.">- <a target="_blank" rel="ugc" href="https://github.com/Blueforcer/ComfyUI-DLSS5-Enhancer">ComfyUI-DLSS5-Enhancer</a> will be included in the next v8.0 update! It is a game changer for image quality. Visit their Github page to get a jump ahead if you put this workflow to good use. For current workflows plug it into Save Image: Out.</h3><h3 id="v8.0-will-also-include-my-own-custom-node-pack-to-ensure-this-workflow-will-continue-to-function-on-my-terms.-i-will-be-setting-up-my-own-github-page-when-it-is-ready.">- v8.0 will also include my own custom node pack to ensure this workflow will continue to function on my terms. I will be setting up my own Github page when it is ready.</h3><h3 id="update-2:">Update 2:</h3><h3 id="i-have-finally-solved-the-spotlighting-problem!-add-this-to-the-beginning-of-a-prompt-then-say-what-your-are-replacing-it-with">I have finally solved the spotlighting problem! Add this to the beginning of a prompt then say what your are replacing it with</h3><pre><code>"Remove all studio, set, self-emissive character, spot, vanity, hero key, and color-grading effects. Replace with ....". </code></pre><h3 id="it-can-probably-be-shortened-but-this-works!">It can probably be shortened but this works!</h3><h3 id="example:">Example:</h3><pre><code>"Remove all studio, set, self-emissive character, spot, vanity, hero key, and color-grading effects. Replace with one frosted window of late daylight bounced off wet tile: soft, wide, diffused, dim, ray-traced practical light only. Color grade: Kodak Portra 400. Static frame. Held pose. Shot on an ARRI Alexa Mini, 50mm Master Prime."</code></pre>10/02/26v7.0 - Inpainting and Outpainting (Standard) Workflow:<img src="https://image.civitai.com/xG1nkqKTMzGDvpLrqFT7WA/c41d89e8-a6dd-4f5f-9dba-4b62f76d15f6/original=true/c41d89e8-a6dd-4f5f-9dba-4b62f76d15f6.jpeg" /><pre><code>v7.0 Inpaint &amp; Outpaint (Standard):
 - Separate workflow.
 - Inpaint Crop And Stitch nodes.
 - Targeted masked editing without having to render the entire image.
 - Extend a image in any direction. Only renders the extension.
 - For the pulled DasiwaMinimaxH3_dasiwaREF2VAHybridV1, DaSiWa MiniMax H3v3 &amp; PinkCherry v5 or v6.
 - Two samplers: Sampler Custom, and Clown Shark Sampler from RES4LYF, swap between them with the Fast Groups Muter.
 - ClownShark group has a chain sampler for the second half of total step count, added a bit more sharpness.
 - Several RES4LYF option nodes: Detail Boost, Implicit Steps, Momentum and Sigma Scaling. Notes from the official RES4LYF example workflow for explanations of what they do are included, along with some quick notes on my testing of them.

Node Requirements:
 - ComfyUI-Inpaint-CropAndStitch
 - ComfyUI-MM-1Frame
 - ComfyUI-Easy-Use
 - ComfyUI-KJNodes
 - RES4LYF
 - rgthree-comfy

v7.0 Updates:
Changed reference image input resolution options. Manual and pass through added, megapixels removed.
 - Separate mute for each reference image, save image and sampler group.
 - Added notes for preliminary Pink Cherry model sampler settings, and input resolution modes.
Ghost link clean.</code></pre>.v7.0 - Single Frame Text-to-Image/Multi-Reference Image-to-Image (Standard) Workflow:<img src="https://image.civitai.com/xG1nkqKTMzGDvpLrqFT7WA/db324827-dd06-47a3-bce3-dedec29bac98/original=true/db324827-dd06-47a3-bce3-dedec29bac98.jpeg" /><pre><code>v7.0 Standard:
 - For the pulled DasiwaMinimaxH3_dasiwaREF2VAHybridV1, DaSiWa MiniMax H3v3 &amp; PinkCherry v5 or v6.
 - Two samplers: Sampler Custom, and Clown Shark Sampler from RES4LYF, swap between them with the Fast Groups Muter.
 - ClownShark group has a chain sampler for the second half of total step count, added a bit more sharpness.
 - Several RES4LYF option nodes: Detail Boost, Implicit Steps, Momentum and Sigma Scaling. Notes from the official RES4LYF example workflow for explanations of what they do are included, along with some quick notes on my testing of them.

Node Requirements:
 - ComfyUI-MM-1Frame
 - ComfyUI-Easy-Use
 - ComfyUI-KJNodes
 - RES4LYF
 - rgthree-comfy

v7 Updates:
 - Changed reference image input resolution options. Manual and pass through added, megapixels removed.
 - Separate mute for each reference image, save image and sampler group.
 - Updated megapixel chart, now goes to 16MP (WOW! 16MP for portraits with V1 model).
 - Added notes for preliminary Pink Cherry model sampler settings, and input resolution modes.
 - Ghost links cleaned.</code></pre><pre><code></code></pre>v7.0 - Inpainting and Outpainting (Basic) Workflow:<img src="https://image.civitai.com/xG1nkqKTMzGDvpLrqFT7WA/3a5ed9d3-ac05-443f-9fe1-f9c295fbca77/original=true/3a5ed9d3-ac05-443f-9fe1-f9c295fbca77.jpeg" /><pre><code>v7.0 - Inpainting and Outpainting (Basic):
 - Separate workflow.
 - Inpaint Crop And Stitch nodes.
 - Targeted masked editing without having to render the entire image.
 - Extend a image in any direction. Only renders the extension.
 - For DaSiWa MiniMax H3v3 &amp; PinkCherry v5 or v6.
 - No RES4LYF custom nodes needed.
 - Has a Beta Sampling Scheduler node using the Beta57 Alpha and Beta's. When switching on Beta57, Denoise and Scheduler are disabled.

Node Requirements:
 - ComfyUI-Inpaint-CropAndStitch
 - ComfyUI-MM-1Frame
 - ComfyUI-Easy-Use
 - ComfyUI-KJNodes
 - rgthree-comfy

v7 Updates:
 - Introduction.
 - Ghost link clean.</code></pre>.v7.0 - Single Frame Text-to-Image/Multi-Reference Image-to-Image (Basic) Workflow:<img src="https://image.civitai.com/xG1nkqKTMzGDvpLrqFT7WA/d61aac98-314e-458b-8821-669e570fa542/original=true/d61aac98-314e-458b-8821-669e570fa542.jpeg" /><pre><code>v7.0 Basic:
 - For DaSiWa MiniMax H3v3 &amp; PinkCherry v5 or v6.
 - No RES4LYF custom nodes needed.
 - Has a Beta Sampling Scheduler node using the Beta57 Alpha and Beta's. When switching on Beta57, Denoise and Scheduler are disabled.

Node Requirements:
 - ComfyUI-MM-1Frame
 - ComfyUI-Easy-Use
 - ComfyUI-KJNodes
 - rgthree-comfy

v7 Updates:
 - Changed reference image input resolution options. Manual and pass through added, megapixels removed.
 - Separate mute for each reference image and save image.
 - Updated megapixel chart, now goes to 16MP.
 - Added notes for preliminary Pink Cherry model sampler settings, and input resolution modes.
 - Ghost links cleaned.
</code></pre><h3 id="the-dasiwaminimaxh3_dasiwaref2vahybridv1-model-darksidewalker-pulled-dasiwa-minimax-h3v3-and-sexgod1979's-pinkcherry-v5-or-v6-are-the-recommended-models-to-use.-dasiwa-if-you-like-smooth-colorful-images-and-pinkcherry-if-you-like-nsfw-raw-and-gritty-images.">The DasiwaMinimaxH3_dasiwaREF2VAHybridV1 model Darksidewalker pulled, DaSiWa MiniMax H3v3 and sexgod1979's PinkCherry v5 or v6 are the recommended models to use. DaSiwa if you like smooth colorful images and PinkCherry if you like NSFW, raw and gritty images.</h3>I tried most of the samplers and landed on what feels like the sweet spot for Clown Shark: diag_implicit/pareschi_russo_2s with Beta57. It’s a little slower, but totally worth it IMO.This workflow is pretty much a set-and-forget workflow, I found settings that give a nice balanced image so nothing "should" need adjusted, still tweak to your hearts content as there is plenty to adjust if you want.The Big Three:The three biggest contributors to image quality is ETA in Sampler Group and Implicit Steps, Lying Strength with Lying Start Step in the Options GroupAdjusting the two down in Lying will produce a more grainy but sharper image, up will make them smooth and blurrier. ETA in Sampler Group does the same thing.Implicit Steps can have multiple effects, sometimes 2 looks better sometimes 1 will, varies from prompt to prompt and seed to seed. Switch to 2 when you get a image you like to save time. 2 is also good for busy scenery.<pre><code></code></pre>Quick notes:<ul><li>The Beta57 scheduler is a must for image quality.</li></ul><ul><li>For extra detail, run a SeedVR detail pass with the same resolution as the image (no upscaling).</li><li>I’d keep the LoRA strength light as these were made for video.</li><li>Pretty much everything you can tweak is sitting on the subgraphs. Fair warning: cranking the resolution too high will probably throw errors.</li><li>Reference Image input size has a great deal to do with rendering time, the bigger the image the slower the renders. If you have a slower card set this lower.</li></ul><pre><code></code></pre>About lighting:<h3 id="i-have-finally-solved-the-spotlighting-problem!-add-this-to-the-beginning-of-a-prompt-then-say-what-your-are-replacing-it-with:">I have finally solved the spotlighting problem! Add this to the beginning of a prompt then say what your are replacing it with:</h3><pre><code> "Remove all studio, set, self-emissive character, spot, vanity, hero key, and color-grading effects. Replace with ....".</code></pre><h3 id="it-can-probably-be-shortened-but-this-works!">It can probably be shortened but this works!</h3><h3 id="example:">Example:</h3><pre><code>"Remove all studio, set, self-emissive character, spot, vanity, hero key, and color-grading effects. Replace with one frosted window of late daylight bounced off wet tile: soft, wide, diffused, dim, ray-traced practical light only. Color grade: Kodak Portra 400. Static frame. Held pose. Shot on an ARRI Alexa Mini, 50mm Master Prime."</code></pre>.Some things I learned on this journey and a ongoing blog. MM in this format loves, loves and loves some more the details. If you don't explain exactly what you want in a detailed way you will get bland output. For example just saying a woman is moaning won't get you much, putting things in like the state of her face such as open mouth, furrowed eyebrows and facial tension will get you much further, this applies to everything, even more when it comes to lighting.<pre><code></code></pre>Nodes and models you will need to have installed, the list below is the bare minimum.custom_nodes:<ul><li><a target="_blank" rel="ugc" href="https://github.com/cicalooo/ComfyUI-MM-1Frame">ComfyUI-MM-1Frame</a> (Basic, Inpaint &amp; Standard)</li><li><a target="_blank" rel="ugc" href="https://github.com/yolain/ComfyUI-Easy-Use">ComfyUI-Easy-Use</a> (Basic, Inpaint &amp; Standard)</li><li><a target="_blank" rel="ugc" href="https://github.com/kijai/ComfyUI-KJNodes">ComfyUI-KJNodes</a> (Basic, Inpaint &amp; Standard)</li><li><a target="_blank" rel="ugc" href="https://github.com/ClownsharkBatwing/RES4LYF">RES4LYF</a> (Inpaint &amp; Standard</li><li><a target="_blank" rel="ugc" href="https://github.com/rgthree/rgthree-comfy">rgthree-comfy</a> (Basic, Inpaint &amp; Standard)</li><li><a target="_blank" rel="ugc" href="https://github.com/lquesada/ComfyUI-Inpaint-CropAndStitch">ComfyUI-Inpaint-CropAndStitch</a> (Inpaint)</li></ul>## Model Linksdiffusion_models:<ul><li><a target="_blank" rel="ugc" href="https://civarchive.com/models/2877206/dasiwa-minimax-h3?modelVersionId=3374445">⛩️💀 DaSiWa MiniMax H3v3 💀⛩️</a></li><li><a target="_blank" rel="ugc" href="https://civarchive.com/models/2838593/pinkcherry-mm-h3-fl2va?modelVersionId=3209732">PinkCherry MM H3 fl2va</a> (Non-V1 favorites! PinkCherry v5 or v6 @10-20 steps, no turbo lora)</li></ul>text_encoders:<ul><li><a target="_blank" rel="ugc" href="https://huggingface.co/Comfy-Org/MiniMax-H3/resolve/main/text_encoders/qwen3vl_32b_minimax_h3_nvfp4_awq.safetensors">qwen3vl_32b_minimax_h3_nvfp4_awq.safetensors</a></li></ul>vae:<ul><li><a target="_blank" rel="ugc" href="https://huggingface.co/Mamad8/MiniMax-H3-Image-VAE/resolve/main/minimax_h3_t1_image_vae_step1597.safetensors">minimax_h3_t1_image_vae_step1597.safetensors</a></li></ul>vae_approx (TAE):<ul><li><a target="_blank" rel="ugc" href="https://huggingface.co/Kijai/MiniMax-H3-TAE/resolve/main/vae_approx/taeh3.safetensors?download=true">taeh3.safetensors (MiniMax H3 TAE)</a></li></ul>loras/h3:<ul><li><a target="_blank" rel="ugc" href="https://huggingface.co/lightx2v/Minimax-h3-Turbo/blob/main/minimax_h3_fl2v_turbo_4step_v1.2_768p_comfyui_bf16.safetensors">minimax_h3_fl2v_turbo_4step_v1.2_768p_comfyui_bf16.safetensors</a></li><li><a target="_blank" rel="ugc" href="https://civarchive.com/models/2929860/minimax-h3-turbo-or-ref2va-or-fl2va-or-hybrid">DaSiWa Minimax H3 Turbo | REF2VA | FL2VA | Hybrid</a> (4.78gb)</li><li>[<a target="_blank" rel="ugc" href="https://civarchive.com/models/2969395/mmh3-afterlight">MMH3] Afterlight</a> (Set @0.17 strength)</li></ul>## Highly Recommendedtext_encoders:<ul><li><a target="_blank" rel="ugc" href="https://huggingface.co/YifeiDevs/Huihui-Qwen3-8B-abliterated-v2-FP8-Comfy/resolve/main/Huihui-Qwen3-8B-abliterated-v2-FP8.safetensors?download=true">Huihui-Qwen3-8B-abliterated-v2-FP8-Comfy</a></li></ul>

MiniMax H3 - Text-to-Image | Multi-Reference Image-to-Image | Inpaint-Outpaint Workflows

Best quality settings:<pre><code>= Standard Workflow =
Resolution: 8-16GP
Model: DasiwaMinimaxH3_dasiwaREF2VAHybridV1
Turo LoRa: minimax_h3_fl2v_turbo_4step_v1.2_768p_comfyui_bf16.safetensors

= ClownShark Settings =
Sampler
 - Steps: 8
 - Sampler: diag_implicit/pareschi_russo_2s
 - Scheduler: Beta57
 - Sampler ETA: 0.0

LoRa
[MMH3] Afterlight
 - Strength: 0.17

Turbo LoRa
 - Strength: 0.5 (Implicit set to: 1)

= Clown Options Settings =
Detail Boost
 - ETA: 4.0
 - Start Step: 1
 - End Step: Keep at either 1/2 of sampler steps or 1 below the final step.

Implicit Steps
 - Implicit Type Sub-steps: 1

Sigma Scaling
 - Noise Anchor SDE: 0.90
 - Lying Sigma: 1.01, Start Step: 1</code></pre>

Files with SHA256: 7b9242bd288b8fd47d853026dbad61eaca813fc8a87b5b595225ead240bfa6cd