═══════════════════════════════════════════════════════════════════════════════ MINIMAX H3 PHOTOREALISTIC IMAGE GENERATOR (LORA + WORKFLOW) ═══════════════════════════════════════════════════════════════════════════════ ─────────────────────────────────────────────────────────────────────────────── IDENTITY ─────────────────────────────────────────────────────────────────────────────── Base model MiniMax H3 Generation mode First-and-Last-Frame-to-Video (FL2VA) Checkpoint not determined — verify before use Category Photorealism Author JonXL Source https://civitai.com/models/2884206 Provenance CivitAI · model 2884206 · version 3260165 ─────────────────────────────────────────────────────────────────────────────── WEIGHTS ─────────────────────────────────────────────────────────────────────────────── File h3_photoreal_ph0t0r34l_1024-step00003780.safetensors Version v1.0 Size 568.8 MB Format SafeTensor / unknown precision SHA256 446f508a699a0dfe53ec97367eb52e802a88643abec6fb1bc63a5a5b93b7747d Published unknown ─────────────────────────────────────────────────────────────────────────────── TRIGGER WORDS ─────────────────────────────────────────────────────────────────────────────── › "ph0t0r34l" ─────────────────────────────────────────────────────────────────────────────── RECOMMENDED SETTINGS ─────────────────────────────────────────────────────────────────────────────── The author did not publish specific settings. Start at strength 0.7 and adjust; see REFERENCE.txt for the MiniMax H3 community baseline. ─────────────────────────────────────────────────────────────────────────────── ADOPTION & POPULARITY ─────────────────────────────────────────────────────────────────────────────── No statistics available. ─────────────────────────────────────────────────────────────────────────────── COMMUNITY NOTES ─────────────────────────────────────────────────────────────────────────────── 27 comment(s) on the source listing; showing 5. • The question is Can it goon • The workflow seems to require an input image? Even when text to video or manually bypassed the loan image node throws and error? [+1] what about image edit? :) • Minor update to workflow 24.08 Workflow version V2 no longer needs an arbitrary image to be loaded for the image generation to take place. [+2] What about ref2va? Archive summary: Minimax H3 Photorealistic Image Generator (LORA + Workflow) is a style LoRA designed for MiniMax H3 checkpoints, published by [JonXL](/users/JonXL) on CivitAI in August 2026. Load a compatible checkpoint first, then apply this LoRA on top via your UI's LoRA loader. It has attracted 2,033 downloads. It's intended for photorealism, image generation, and h3 image generator content. Source: https://civarchive.com/models/2884206?modelVersionId=3260165 Comments are other users' words, reproduced as written. ─────────────────────────────────────────────────────────────────────────────── CHAINING / STACKING ─────────────────────────────────────────────────────────────────────────────── Stacks with others yes Strength when stacked Use a LOWER LoRA strength when running it together with your own character LoRAs. For the "combining both loras" experiment the author only says to try increasing/decreasing strength and running them at different strengths — no specific stacked values given. What the author says (fact-checked against their own text): Author states the LoRA can be run alongside the user's own character LoRAs at lower strength: "The lora when run at lower lora strength can also help bring out any of your character loras n a more static position." He also floats, as an experiment, "combining both loras with different strengths" — but he never says which two LoRAs "both loras" refers to, so that is left unspecified. Nothing is stated about load or chain order. Requires alongside it: ! Cicaloo 1 Frame custom node plus its special VAE — required when using the Cicaloo 1 Frame workflows: "make sure to clome the custom node and download the special VAE, no turbo lora required" (node repo: https://github.com/cicalooo/ComfyUI-MM-1Frame). No turbo/speed LoRA is needed. ! Kijai's nodes (ComfyUI-KJNodes) and VideoHelperSuite — hedged: "You may need to install some frequently used nodes, mostly Kijai's nodes and VideoHelperSuite" Do NOT combine with: x Reference model + the Cicaloo 1 Frame workflow — the author states reference-model support exists in the default workflow only: "With support for reference model in the default workflow (not the 1 Frame workflow)" General MiniMax H3 rules that apply here (see REFERENCE.txt): • Load the LoRA with LoraLoaderModelOnly between UNETLoader and your sampler; stacking means chaining several of those nodes in series. • Stacks normally on top of a single acceleration/turbo LoRA; keep only one distill LoRA active in the chain. • The author never states which checkpoint family this targets. Test against your own workflow before relying on it, and remember a mismatch degrades output silently rather than erroring. ─────────────────────────────────────────────────────────────────────────────── PROS ─────────────────────────────────────────────────────────────────────────────── + Well established — 2,024 downloads. [source: CivitAI stats] + Positively received — 95 thumbs-up with no down-votes. [source: CivitAI stats] + Explicit trigger word(s) — ph0t0r34l — so the effect can be turned on and off from the prompt. [source: model metadata] ─────────────────────────────────────────────────────────────────────────────── CONS & CAVEATS ─────────────────────────────────────────────────────────────────────────────── - Thin documentation — expect trial and error. [source: description length] ─────────────────────────────────────────────────────────────────────────────── COMPATIBILITY WARNING ─────────────────────────────────────────────────────────────────────────────── MiniMax H3 ships two transformers that share tensor key names: the fl2va checkpoint (driving T2VA, I2VA, L2VA and FL2VA) and the separate ref2va checkpoint (driving Ref2VA). Loading a LoRA against the wrong family can silently degrade output instead of raising an error, so confirm the family above matches the checkpoint you have loaded. Likewise, LoRAs built for full checkpoints often throw adaln_proj shape errors on pruned ones — match the build to your checkpoint. ─────────────────────────────────────────────────────────────────────────────── AUTHOR'S DESCRIPTION ─────────────────────────────────────────────────────────────────────────────── (the author left no description) ─────────────────────────────────────────────────────────────────────────────── EXAMPLE MEDIA ─────────────────────────────────────────────────────────────────────────────── 6 preview(s) in ./example_images/ — community results for this LoRA. Videos keep a matching _poster.jpg still frame. No prompts are recorded below: CivitAI's public API flags these posts as having generation metadata but does not return it, so the prompts cannot be scraped. Open the model page to read them. 01_video.mp4 (video, 1376x802) still: 01_poster.jpg 02_image.jpeg (image, 2560x1440) 03_image.jpeg (image, 2560x1440) 04_image.jpeg (image, 2560x1440) 05_image.jpeg (image, 2560x1440) 06_image.jpeg (image, 2560x1440) ─────────────────────────────────────────────────────────────────────────────── HOW TO USE (COMFYUI) ─────────────────────────────────────────────────────────────────────────────── 1. Copy the .safetensors file into ComfyUI/models/loras/ 2. Load the matching H3 checkpoint with UNETLoader. 3. Insert LoraLoaderModelOnly between the loader and your sampler. 4. Set strength as listed above, then add any trigger word to the prompt. 5. Keep CFG at 1.0 with BasicGuider unless you know you want real CFG. ─────────────────────────────────────────────────────────────────────────────── Scraped by lorakit on 2026-09-05 03:47 UTC. Stats and description are the author's; pros/cons are derived from the evidence tagged beside each line. ───────────────────────────────────────────────────────────────────────────────