# Qwen Image 2.1 (engine `qwen21`) — host weights + Comfy graph

## Host paths (this RTX 5080 box)

| Role | Resolved path |
|---|---|
| ComfyUI root (Klein / host agent `ComfyUI (1)`) | `C:\Users\ianjm\AppData\Local\Comfy-Desktop\ComfyUI-Installs\ComfyUI (1)\ComfyUI` |
| Models root (Desktop Shared via `shared_model_paths.yaml`) | `C:\Users\ianjm\AppData\Local\Comfy-Desktop\ComfyUI-Shared\models` |

Weights must land under Shared so every Desktop instance sees them:

- `diffusion_models\qwen-image-2.1-Q8_0.gguf` (~7.59 GiB) — from `abenzerps/Qwen-Image-2.1-Uncensored-GGUF`
- `text_encoders\qwen3vl_8b_int8_convrot.safetensors` — from `Comfy-Org/Qwen-Image-2.1` (INT8; BF16 8B VL OOMs on 16 GB)
- `vae\qwen_image_2.1_vae_bf16.safetensors` — from `Comfy-Org/Qwen-Image-2.1` (**not** the old Qwen-Image 1.0 VAE)

Custom node: `custom_nodes\ComfyUI-GGUF` (`UnetLoaderGGUF`). Do not install a second GGUF pack.

## Setup

```powershell
powershell -ExecutionPolicy Bypass -File scripts\setup-qwen21.ps1
```

DiT download uses the exact host invocation:

```powershell
hf download hf://abenzerps/Qwen-Image-2.1-Uncensored-GGUF/qwen-image-2.1-Q8_0.gguf
```

That lands in the Hugging Face hub cache; the setup script copies it into Shared `diffusion_models`. The abenzerps GGUF ships with `kv_count=0` (no `general.architecture`) **and** Q8_0-quantized 1D RMSNorm weights (logical 128 → packed 136), which breaks Comfy’s fused `rms_rope`. `scripts/tag-qwen21-gguf.py` rewrites the file with `general.architecture=qwen_image` and promotes small/1D tensors to F32 so city96 `UnetLoaderGGUF` can load it.

Restart Comfy **only when idle** (`COMFY_CONTROL_URL/status` → `gpu.busy=false`). Confirm `object_info` lists `UnetLoaderGGUF` and the three filenames.

## App

- Engine key: `qwen21` · UI label: **Qwen 2.1**
- **Generate** (T2I) and **Edit** (same checkpoint, second graph). Compose / Iterate / Video / Extend / Music stay disabled unless a graph exists.
- Edit requires Start still → `images.image_1`; prompt slots use `<image1>` / `<image2>`. If the user omits `<image1>`, the runner prepends `Keep the identity, face, body, and pose of the person in <image1>. Apply only this change:` then their text. Hero is Klein-only and is ignored (never image_1).
- Face / outfit locks stay Klein semantics — they do not drive Qwen slots.
- Sampler defaults: euler / simple / cfg **1** / steps **25**. Edit uses `QwenImage21Cache` (device auto, dtype int8). T2I uses `ModelSamplingAuraFlow` shift **3.1**.
- Default Generate canvas follows the bench Aspect control on the Qwen-safe ~1 MP grid (`EmptyLatentImage`): 1:1 → 1024×1024, 16:9 → 1536×864, 9:16 → 864×1536 (and the other table rows). PE `wh_ratio` is advisory only and never sizes the canvas. Do not use native 2K bins on 16 GB.
- Edit follows `image_1` via the encode node’s latent (resolution long-edge ~1024). Do not inject a picker EmptyLatentImage onto the edit sampler.
- Optional **Enhance prompt** (off by default): runs a separate PE-only Comfy graph (`CLIPLoader` + rewrite node), then frees VRAM and queues the existing T2I/Edit graph with the rewritten prompt. Never loads PE CLIP + DiT together on 16 GB. Fail closed if `parse_ok` is false or the rewrite is empty.
  - Generate + Enhance → PE-T2I only (`pe_t2i`). No images.
  - Edit + Enhance → PE-I2I only (`pe_i2i`) with the same Start still on `image_1`. After PE, the sample prompt is always the keep-identity stanza + typed instruction first (`<image1>` in front). A PE rewrite is appended only when it is an edit directive; T2I-style observer captions (`The image is…`) are dropped (`enhance.skippedAsDescribe`) and the stanza + raw remain. Never let the PE chunk be the entire prompt. Library stores `prompt` (TextEncode string) and `promptRaw` (Typed). Never fall back to PE-T2I on Edit.
  - If PE refuses or returns an empty/gutted rewrite, the job continues with `promptRaw` (Edit still gets the keep-identity stanza). Details shows `Enhance skipped (model refused) — used your prompt.` (`enhance.refused`).
  - Host PE system prompts live in `host/qwen21-pe-prompts/` and are copied onto the node pack by `scripts/setup-qwen21.ps1` (official steps + Adult appendix). Edit prompt must lead with an operation and `<image1>` — never “The image is a photograph of…”. Do not swap Heretic PE weights on 16 GB.
- PE weights (int8 only): `text_encoders\qwen3.5_9b_qwen_image_2.1_pe_{t2i,i2i}.int8_convrot.safetensors` — do not replace the image TE `qwen3vl_8b_int8_convrot.safetensors`.
- Optional **Turbo** (off by default): Viggle DMD LoRA `Qwen-Image-2.1-viggle-turbo-v0.2.1-6step-lora-r128.safetensors` on the same Q8 GGUF via `ViggleTurboLora` + `ViggleTurboSigmas` (`1.0, 0.9375, 0.875, 0.75, 0.5, 0.25`), 6 steps, CFG 1, empty negative. Not a new engine. Enhance prompt stays compatible and is recommended with Turbo.
- Custom node: `ComfyUI-Viggle-Turbo` (`viggle_turbo.py`). Do not merge the LoRA with stock `LoraLoaderModelOnly` (lossy on int8/bf16).

Graphs: `server/assets/studio2_qwen21_t2i.json`, `server/assets/studio2_qwen21_edit.json`, `server/assets/studio2_qwen21_t2i_turbo.json`, `server/assets/studio2_qwen21_edit_turbo.json`, `server/assets/studio2_qwen21_pe_t2i.json`, `server/assets/studio2_qwen21_pe_edit.json`.