Add low-vram workflow batch for Z-Image, LTX, and MiniMax

models/ltx/ltx23-i2v-t2v-basic-checkpoint-models/README.md

@@ -0,0 +1,43 @@
+# LTX 2.3 Basic for Checkpoint Models
+
+## Summary
+
+`LTX 2.3` basic workflow variant intended for checkpoint-model loading rather than only split-model wiring. Useful if you want a more standard loading path while still staying inside the RuneXX ecosystem.
+
+## Model Family
+
+- `ltx`
+
+## Status
+
+- `imported-only`
+
+## Source
+
+- Workflow pack page: <https://huggingface.co/RuneXX/LTX-2.3-Workflows>
+- Workflow file: <https://huggingface.co/RuneXX/LTX-2.3-Workflows/resolve/main/LTX-2.3_-_I2V_T2V_Basic_for_checkpoint_models.json?download=true>
+- Supporting split/checkpoint model page: <https://huggingface.co/Kijai/LTX2.3_comfy>
+- Original publisher: `RuneXX`
+- Date imported: `2026-08-08`
+
+## Developer notes
+
+- RuneXX says the extracted-model path may run easier, but also notes you can swap to default Comfy checkpoint loading if desired.
+- This workflow includes more tuning and helper nodes than the simple single-pass graphs.
+
+## Our notes
+
+- More flexible than the simplest LTX imports, but also more dependency-heavy.
+- On a `3070`, this is more of a “second test” workflow after the single-pass variants.
+- Worth keeping because it may be a better base for later cleanup if the simpler workflows are too stripped down.
+
+## Required custom nodes
+
+- `ComfyUI-KJNodes`
+- `rgthree-comfy`
+- `easy use` / `easy showAnything`
+- optional `ComfyUI-GGUF` hooks are present
+
+## Notes
+
+- Better treated as a richer base graph than as the first low-VRAM test.

models/ltx/ltx23-i2v-t2v-simple-single-pass/README.md

@@ -0,0 +1,45 @@
+# LTX 2.3 I2V/T2V Simple Single Pass
+
+## Summary
+
+Community `LTX 2.3` single-pass workflow supporting both image-to-video and text-to-video. Chosen as a lower-complexity starting point for tighter VRAM situations.
+
+## Model Family
+
+- `ltx`
+
+## Status
+
+- `imported-only`
+
+## Source
+
+- Workflow pack page: <https://huggingface.co/RuneXX/LTX-2.3-Workflows>
+- Workflow file: <https://huggingface.co/RuneXX/LTX-2.3-Workflows/resolve/main/LTX-2.3_-_I2V_T2V_Simple_single_pass.json?download=true>
+- Supporting split-model page: <https://huggingface.co/Kijai/LTX2.3_comfy>
+- Original publisher: `RuneXX`
+- Date imported: `2026-08-08`
+
+## Developer notes
+
+- RuneXX says the workflows are based on extracted models from `Kijai/LTX2.3_comfy`.
+- The repo explicitly says the extracted models may run easier on your computer as separate files.
+- The README also recommends keeping `ComfyUI-KJNodes`, `ComfyUI-GGUF`, and ComfyUI itself fully up to date.
+
+## Our notes
+
+- This is one of the better `3070` candidates in the LTX set because it aims for a simpler single-pass route rather than the larger story/multi-scene graphs.
+- Still not truly light, but it is more realistic than the giant movie-maker style workflows.
+- Good first LTX import when you want to see whether the family is practical at all on your box.
+
+## Required custom nodes
+
+- `ComfyUI-KJNodes`
+- `rgthree-comfy`
+- `easy use` / `easy showAnything`
+- optional `ComfyUI-GGUF` support is present in the graph, though not required if you stay on standard loaders
+
+## Notes
+
+- Best LTX first try for low-complexity local testing.
+- If this feels too heavy, the next thing to optimize is usually preview/decode choices and model-loading path.

models/ltx/ltx23-t2v-simple-single-pass/README.md

@@ -0,0 +1,42 @@
+# LTX 2.3 T2V Simple Single Pass
+
+## Summary
+
+Simpler text-to-video only `LTX 2.3` workflow from the RuneXX pack. Chosen because it avoids some of the extra image-input complexity while still staying in the lower-bloat part of the workflow pack.
+
+## Model Family
+
+- `ltx`
+
+## Status
+
+- `imported-only`
+
+## Source
+
+- Workflow pack page: <https://huggingface.co/RuneXX/LTX-2.3-Workflows>
+- Workflow file: <https://huggingface.co/RuneXX/LTX-2.3-Workflows/resolve/main/LTX-2.3_-_T2V_Simple_single_pass.json?download=true>
+- Supporting split-model page: <https://huggingface.co/Kijai/LTX2.3_comfy>
+- Original publisher: `RuneXX`
+- Date imported: `2026-08-08`
+
+## Developer notes
+
+- Same general workflow family as the other RuneXX imports, built around the extracted `Kijai` model layout.
+- The pack is designed to support both standard and GGUF-oriented loading choices, depending on what the user has installed.
+
+## Our notes
+
+- This is probably the simplest `LTX 2.3` text-to-video import we currently have.
+- On a `3070`, it is still a “tight to heavy” experiment, but it is a better first look than the more elaborate prompt-relay or long-video graphs.
+- Good candidate for a later optimized low-complexity LTX master if the outputs are decent.
+
+## Required custom nodes
+
+- `ComfyUI-KJNodes`
+- `rgthree-comfy`
+- `easy use` / `easy showAnything`
+
+## Notes
+
+- Best LTX pure T2V starter in the current batch.

models/minimax/minimax-h3-official-i2v/README.md

@@ -0,0 +1,40 @@
+# MiniMax H3 Official I2V
+
+## Summary
+
+Official `MiniMax H3` image-to-video template from the Comfy workflow templates. Good baseline for seeing how the official Comfy packaging expects MiniMax to be wired.
+
+## Model Family
+
+- `minimax`
+
+## Status
+
+- `imported-only`
+
+## Source
+
+- Model card: <https://huggingface.co/Comfy-Org/MiniMax-H3>
+- Workflow file: <https://raw.githubusercontent.com/Comfy-Org/workflow_templates/refs/heads/main/templates/video_minimax_h3_i2v.json>
+- Original publisher: `Comfy-Org`
+- Date imported: `2026-08-08`
+
+## Developer notes
+
+- The model card lists several pruned and scaled model variants, including `pruned_fp8_scaled` files, which are the most interesting ones for tighter hardware.
+- The official card also separates diffusion models, text encoders, and VAEs clearly, which is useful when wiring custom local setups.
+
+## Our notes
+
+- This is the safest MiniMax baseline we have.
+- On a `3070`, MiniMax still looks tight to heavy, but the pruned/fp8-scaled model variants make it worth keeping on the board.
+- If MiniMax becomes practical for you, this official workflow is the best starting point before moving into community turbo variants.
+
+## Required custom nodes
+
+- current official Comfy support for MiniMax templates
+
+## Notes
+
+- Best first official MiniMax import.
+- Most relevant low-VRAM angle is choosing the pruned/fp8-scaled files from the model card rather than the heavier variants.

models/minimax/minimax-h3-official-t2v/README.md

@@ -0,0 +1,38 @@
+# MiniMax H3 Official T2V
+
+## Summary
+
+Official `MiniMax H3` text-to-video template from the Comfy workflow templates. Useful for direct prompt-to-video testing without needing an image input.
+
+## Model Family
+
+- `minimax`
+
+## Status
+
+- `imported-only`
+
+## Source
+
+- Model card: <https://huggingface.co/Comfy-Org/MiniMax-H3>
+- Workflow file: <https://raw.githubusercontent.com/Comfy-Org/workflow_templates/refs/heads/main/templates/video_minimax_h3_t2v.json>
+- Original publisher: `Comfy-Org`
+- Date imported: `2026-08-08`
+
+## Developer notes
+
+- The official card links this directly as one of the three primary workflows (`I2V`, `T2V`, `R2V`).
+- Same model-card storage guidance applies here, including the pruned and scaled diffusion-model variants.
+
+## Our notes
+
+- This is the cleanest MiniMax T2V baseline available from an official source.
+- On a `3070`, this is still a reach, but it is worth keeping as the control case when comparing against the turbo LoRA path.
+
+## Required custom nodes
+
+- current official Comfy support for MiniMax templates
+
+## Notes
+
+- Use this as the baseline when deciding whether the community turbo workflow is actually buying enough speed.

models/minimax/minimax-h3-turbo-lora-community/README.md

@@ -0,0 +1,51 @@
+# MiniMax H3 Turbo LoRA Community Workflow
+
+## Summary
+
+Community `MiniMax H3` turbo LoRA workflow intended to cut step count and generation time dramatically compared with the standard MiniMax path.
+
+## Model Family
+
+- `minimax`
+
+## Status
+
+- `imported-only`
+
+## Source
+
+- Community repo page: <https://huggingface.co/drbaph/MiniMax-H3-Turbo-Lora-ComfyUI>
+- Workflow file: <https://huggingface.co/drbaph/MiniMax-H3-Turbo-Lora-ComfyUI/resolve/main/fl_minimax_h3_turbo_lora_example_workflow.json?download=true>
+- Original LoRA project credited by the maintainer: <https://huggingface.co/larryvrh/MiniMax-H3-Turbo-Lora>
+- Original publisher of compatibility conversion: `drbaph`
+- Date imported: `2026-08-08`
+
+## Developer notes
+
+- The maintainer says this repo contains ComfyUI compatibility conversions of the original Turbo LoRA so it works with the pruned/curve-form MiniMax model in ComfyUI.
+- Recommended checkpoint: `minimax_h3_turbo_v4_step600_ema_pruned_comfyui.safetensors`
+- Recommended settings:
+  - `8` steps
+  - `euler`
+  - `beta`
+  - LoRA strength `1.0`
+  - Video sigma shift `12`
+  - Audio sigma shift `4-6`
+- The repo explicitly positions this as a much lower-step route than standard MiniMax generation.
+
+## Our notes
+
+- This is the most interesting MiniMax import for a `3070` because low-step operation is exactly the kind of optimization you need.
+- Still likely tight, but much more realistic than brute-forcing full standard MiniMax settings.
+- This is the first MiniMax workflow I would actually try for local feasibility after the official baseline is understood.
+
+## Required custom nodes
+
+- MiniMax-compatible Comfy support
+- `ComfyUI-KJNodes`-style preview/attention helpers appear in the graph
+- optional acceleration hooks referenced in the source include SageAttention, Sol Attention, Gradient, and Spectrum
+
+## Notes
+
+- Best MiniMax low-step candidate in the current batch.
+- Strong contender for a future `optimized/text-to-video/` or `optimized/image-to-video/` MiniMax master if it performs well enough on your machine.

models/zimage/README.md

@@ -0,0 +1,12 @@
+# Z-Image Workflows
+
+Use this folder for `Z-Image` and `Z-Image Turbo` workflows.
+
+Recommended focus areas:
+
+- fast text-to-image
+- image-edit variants
+- refinement / first-pass ideation
+- workflows that look better or simpler than comparable FLUX image paths
+
+Store each workflow in its own subfolder.

models/zimage/z-image-turbo-official-example/README.md

@@ -0,0 +1,51 @@
+# Z-Image Turbo Official Example
+
+## Summary
+
+Official `Z-Image Turbo` example workflow distributed as a PNG with embedded Comfy metadata. Good starting point for fast local image generation, especially when you want a simpler image stack than heavier FLUX setups.
+
+## Model Family
+
+- `zimage`
+
+## Status
+
+- `imported-only`
+
+## Source
+
+- Example page: <https://comfyanonymous.github.io/ComfyUI_examples/z_image/>
+- Workflow asset: <https://raw.githubusercontent.com/comfyanonymous/ComfyUI_examples/master/z_image/z_image_turbo_example.png>
+- Model card pointing to the example: <https://huggingface.co/Comfy-Org/z_image_turbo>
+- Original publisher: `ComfyUI_examples` / `Comfy-Org`
+- Date imported: `2026-08-08`
+
+## Developer notes
+
+- The example page describes `Z-Image Turbo` as a fast distilled diffusion model.
+- The workflow is distributed as an image with embedded Comfy metadata, so you load it by dragging the PNG into ComfyUI.
+- The example page points to:
+  - `qwen_3_4b.safetensors`
+  - `z_image_turbo_bf16.safetensors`
+  - `ae.safetensors`
+
+## Our notes
+
+- This is the cleanest official `Z-Image Turbo` starting point we have right now.
+- On a `3070`, this looks more plausible as a practical fast image-generation workflow than many heavier FLUX variants.
+- If `Z-Image` turns into a real favorite, this workflow will likely become one of the main inputs for a future `optimized/text-to-image/` master.
+
+## Required custom nodes
+
+- none beyond current official Comfy support for the embedded example workflow
+
+## Required model files
+
+- `models/text_encoders/qwen_3_4b.safetensors`
+- `models/diffusion_models/z_image_turbo_bf16.safetensors`
+- `models/vae/ae.safetensors`
+
+## Notes
+
+- Load by dragging `workflow.png` into ComfyUI.
+- Best current `Z-Image Turbo` baseline in the repo.