README.md
23.1 KB · 302 lines · markdown Raw
1 ---
2 base_model: Qwen/Qwen-Image-2.1-Turbo
3 base_model_relation: quantized
4 license: other
5 license_name: qwen-research
6 license_link: https://huggingface.co/Qwen/Qwen-Image-2.1-Turbo/blob/main/LICENSE
7 language:
8 - en
9 - zh
10 pipeline_tag: text-to-image
11 library_name: gguf
12 tags:
13 - gguf
14 - quantized
15 - unsloth
16 - qwen
17 - image-generation
18 - turbo
19 widget:
20 - text: a neon-lit ramen shop in a rainy Tokyo alley at night, reflections on wet pavement, cinematic
21 output:
22 url: assets/ramen.png
23 - text: macro photograph of a honeybee on a sunflower, visible pollen grains, extreme detail
24 output:
25 url: assets/honeybee.png
26 ---
27 # Read our How to [Run Qwen-Image-2.1 Guide!](https://unsloth.ai/docs/models/qwen-image-2.1)
28
29 This is a GGUF quantized version of [Qwen-Image-2.1-Turbo](https://huggingface.co/Qwen/Qwen-Image-2.1-Turbo), the 8-step distilled checkpoint of [Qwen-Image-2.1](https://huggingface.co/Qwen/Qwen-Image-2.1). <br>
30 unsloth/Qwen-Image-2.1-Turbo-GGUF uses [Unsloth Dynamic 2.0](https://docs.unsloth.ai/basics/unsloth-dynamic-2.0-ggufs) methodology for SOTA performance.
31
32 - Important layers are upcasted to higher precision, per tensor, from a measured sensitivity scan. Every file uses the same per-tensor layout as the matching file in [unsloth/Qwen-Image-2.1-GGUF](https://huggingface.co/unsloth/Qwen-Image-2.1-GGUF), with an importance matrix collected on Turbo itself at its own 8-step schedule.
33 - Run these with [Unsloth Desktop](https://github.com/unslothai/unsloth), ComfyUI (ComfyUI-GGUF), stable-diffusion.cpp and more. A GGUF is the denoiser only, so it needs the VAE and the Qwen3-VL text encoder alongside it. Turbo uses the same text encoder and VAE as Qwen-Image-2.1.
34 - VAE: [unsloth/Qwen-Image-2.1-FP8](https://huggingface.co/unsloth/Qwen-Image-2.1-FP8) `vae/qwen_image_2.1_vae_bf16.safetensors`. Text encoder: [unsloth/Qwen3-VL-8B-Instruct-GGUF](https://huggingface.co/unsloth/Qwen3-VL-8B-Instruct-GGUF) `Qwen3-VL-8B-Instruct-UD-Q4_K_XL.gguf`, the Dynamic 2.0 4-bit rung rather than the uniform `Q4_K_M`.
35 - Turbo is meant for **8 steps at CFG 1** with the sampling schedule it ships (`sample_sigmas` in [model_index.json](https://huggingface.co/Qwen/Qwen-Image-2.1-Turbo/blob/main/model_index.json)). Unsloth Desktop applies it automatically.
36 <div>
37 <div style="display: flex; gap: 5px; align-items: center; ">
38 <a href="https://github.com/unslothai/unsloth/">
39 <img src="https://github.com/unslothai/unsloth/raw/main/images/unsloth%20new%20logo.png" width="133">
40 </a>
41 <a href="https://discord.gg/unsloth">
42 <img src="https://github.com/unslothai/unsloth/raw/main/images/Discord%20button.png" width="173">
43 </a>
44 <a href="https://unsloth.ai/docs/models/qwen-image-2.1">
45 <img src="https://raw.githubusercontent.com/unslothai/unsloth/refs/heads/main/images/documentation%20green%20button.png" width="143">
46 </a>
47 </div>
48 </div>
49
50 ## Files
51
52 LPIPS (AlexNet, lower is better) against the bf16 Turbo transformer, same text encoder, VAE, prompts and seeds, 1024x1024, 8 steps, CFG 1, rendered through Unsloth Desktop's diffusers route. Mean with 95% bootstrap CI over 8 prompts x 2 seeds; the held-out column uses 8 further prompts x 2 seeds that were not part of the importance-matrix calibration. Two bf16 runs of the same seeds give 0.000 (run-to-run floor).
53
54 | File | Size (GB) | LPIPS | LPIPS (held-out) |
55 |---|---|---|---|
56 | [qwen-image-2.1-turbo-Q2_K.gguf](https://huggingface.co/unsloth/Qwen-Image-2.1-Turbo-GGUF/blob/main/qwen-image-2.1-turbo-Q2_K.gguf) | 2.47 | 0.374 [0.300, 0.453] | 0.351 [0.297, 0.403] |
57 | [qwen-image-2.1-turbo-Q3_K_S.gguf](https://huggingface.co/unsloth/Qwen-Image-2.1-Turbo-GGUF/blob/main/qwen-image-2.1-turbo-Q3_K_S.gguf) | 2.72 | 0.303 [0.229, 0.381] | 0.280 [0.225, 0.334] |
58 | [qwen-image-2.1-turbo-Q3_K_M.gguf](https://huggingface.co/unsloth/Qwen-Image-2.1-Turbo-GGUF/blob/main/qwen-image-2.1-turbo-Q3_K_M.gguf) | 3.17 | 0.261 [0.189, 0.342] | 0.250 [0.193, 0.323] |
59 | [qwen-image-2.1-turbo-Q3_K_XL.gguf](https://huggingface.co/unsloth/Qwen-Image-2.1-Turbo-GGUF/blob/main/qwen-image-2.1-turbo-Q3_K_XL.gguf) | 3.61 | 0.237 [0.148, 0.333] | 0.208 [0.165, 0.256] |
60 | [qwen-image-2.1-turbo-Q4_K_S.gguf](https://huggingface.co/unsloth/Qwen-Image-2.1-Turbo-GGUF/blob/main/qwen-image-2.1-turbo-Q4_K_S.gguf) | 3.91 | 0.201 [0.126, 0.279] | 0.186 [0.138, 0.244] |
61 | [qwen-image-2.1-turbo-Q4_K_M.gguf](https://huggingface.co/unsloth/Qwen-Image-2.1-Turbo-GGUF/blob/main/qwen-image-2.1-turbo-Q4_K_M.gguf) | 4.20 | 0.153 [0.097, 0.215] | 0.141 [0.105, 0.182] |
62 | [qwen-image-2.1-turbo-Q5_K_S.gguf](https://huggingface.co/unsloth/Qwen-Image-2.1-Turbo-GGUF/blob/main/qwen-image-2.1-turbo-Q5_K_S.gguf) | 4.50 | 0.103 [0.064, 0.145] | 0.083 [0.062, 0.110] |
63 | [qwen-image-2.1-turbo-Q5_K_M.gguf](https://huggingface.co/unsloth/Qwen-Image-2.1-Turbo-GGUF/blob/main/qwen-image-2.1-turbo-Q5_K_M.gguf) | 5.39 | 0.074 [0.037, 0.117] | 0.087 [0.049, 0.133] |
64 | [qwen-image-2.1-turbo-Q6_K.gguf](https://huggingface.co/unsloth/Qwen-Image-2.1-Turbo-GGUF/blob/main/qwen-image-2.1-turbo-Q6_K.gguf) | 6.27 | 0.055 [0.028, 0.089] | 0.076 [0.038, 0.121] |
65 | [qwen-image-2.1-turbo-Q6_K_XL.gguf](https://huggingface.co/unsloth/Qwen-Image-2.1-Turbo-GGUF/blob/main/qwen-image-2.1-turbo-Q6_K_XL.gguf) | 6.72 | 0.050 [0.024, 0.083] | 0.069 [0.034, 0.115] |
66 | [qwen-image-2.1-turbo-Q8_0.gguf](https://huggingface.co/unsloth/Qwen-Image-2.1-Turbo-GGUF/blob/main/qwen-image-2.1-turbo-Q8_0.gguf) | 7.64 | 0.036 [0.017, 0.059] | 0.030 [0.016, 0.047] |
67 | [qwen-image-2.1-turbo-F16.gguf](https://huggingface.co/unsloth/Qwen-Image-2.1-Turbo-GGUF/blob/main/qwen-image-2.1-turbo-F16.gguf) | 14.23 | 0.012 [0.005, 0.019] | 0.016 [0.008, 0.025] |
68
69 **Recommended:** `Q4_K_M` (4.20 GB) for most GPUs. For output closest to bf16 use `Q8_0` (7.64 GB); `Q6_K_XL` / `Q6_K` sit in between. The 2- and 3-bit files fit very small cards but drift visibly from bf16 at 8 steps, so prefer `Q3_K_XL` over `Q3_K_M` / `Q3_K_S` / `Q2_K` when you must go below 4 GB.
70
71 ## Run in Unsloth Desktop
72
73 Load a file from `unsloth/Qwen-Image-2.1-Turbo-GGUF` (or a local `qwen-image-2.1-turbo-*.gguf`) for image generation. Unsloth Desktop recognises the Turbo files by name, uses the base `Qwen/Qwen-Image-2.1-Turbo` for the text encoder and VAE, and applies the 8-step schedule with CFG 1.
74
75 ## Run in ComfyUI
76
77 Install [ComfyUI-GGUF](https://github.com/city96/ComfyUI-GGUF), put a `qwen-image-2.1-turbo-*.gguf` file in `models/diffusion_models` (or `models/unet`), `qwen3vl_8b_bf16.safetensors` in `models/text_encoders` and `qwen_image_2.1_vae_bf16.safetensors` in `models/vae` (both from [Comfy-Org/Qwen-Image-2.1](https://huggingface.co/Comfy-Org/Qwen-Image-2.1)). Start from the Qwen-Image-2.1 text-to-image template and replace the `Load Diffusion Model` node with `Unet Loader (GGUF)`. Set the KSampler to **8 steps, CFG 1**, sampler `euler`, scheduler `simple`.
78
79 ## Run with stable-diffusion.cpp
80
81 Pass Turbo's 8-step schedule (plus the terminal 0) as custom sigmas:
82
83 ```bash
84 sd-cli --diffusion-model qwen-image-2.1-turbo-Q4_K_M.gguf \
85 --vae qwen_image_2.1_vae_bf16.safetensors \
86 --llm Qwen3-VL-8B-Instruct-UD-Q4_K_XL.gguf \
87 -p "a cartoon sloth mascot waving, flat vector illustration, bright colours" \
88 --steps 8 --cfg-scale 1.0 --sampling-method euler -W 1024 -H 1024 --diffusion-fa \
89 --sigmas "1.0,0.978453,0.95418,0.926626,0.89508,0.845148,0.704534,0.414568,0.0" \
90 -o out.png
91 ```
92
93 ### Samples
94
95 Rendered with the Q4_K_M denoiser, 1024x1024, 8 steps, CFG 1.
96
97 <table>
98 <tr>
99 <td><img src="https://huggingface.co/unsloth/Qwen-Image-2.1-Turbo-GGUF/resolve/main/assets/ramen.png" width="200"></td>
100 <td><img src="https://huggingface.co/unsloth/Qwen-Image-2.1-Turbo-GGUF/resolve/main/assets/honeybee.png" width="200"></td>
101 </tr>
102 <tr>
103 <td><img src="https://huggingface.co/unsloth/Qwen-Image-2.1-Turbo-GGUF/resolve/main/assets/chalkboard.png" width="200"></td>
104 <td><img src="https://huggingface.co/unsloth/Qwen-Image-2.1-Turbo-GGUF/resolve/main/assets/knight.png" width="200"></td>
105 </tr>
106 </table>
107
108 ---
109 <p align="center">
110 <img src="https://qianwen-res.oss-cn-beijing.aliyuncs.com/Qwen-Image/image2.1/logo.png" width="400"/>
111 </p>
112 <p align="center">
113 <a href="https://modelscope.cn/models/Qwen/Qwen-Image-2.1-Turbo">ModelScope</a>&nbsp;&nbsp;|
114 &nbsp;&nbsp;<a href="https://huggingface.co/Qwen/Qwen-Image-2.1-Turbo">HuggingFace</a>&nbsp;&nbsp;|
115 &nbsp;&nbsp;<a href="https://github.com/QwenLM/Qwen-Image-2.1">GitHub</a>&nbsp;&nbsp;|
116 &nbsp;&nbsp;<a href="https://qwen.ai/blog?id=qwen-image-2.1">Qwen-Image-2.1 Blog</a>&nbsp;&nbsp;|
117 &nbsp;&nbsp;<a href="https://discord.gg/BEYSk3pkSu">Discord</a>&nbsp;&nbsp;|
118 &nbsp;&nbsp;<a href="https://huggingface.co/Qwen/Qwen-Image-2.1/blob/main/assets/qr.png">WeChat</a>
119 </p>
120
121 ## Introduction
122
123 **Qwen-Image-2.1-Turbo** is an accelerated checkpoint of [Qwen-Image-2.1](https://huggingface.co/Qwen/Qwen-Image-2.1) for text-to-image generation and image editing with **8 denoising steps**. It uses the same 7B visual generation architecture and loads directly with `QwenImage21Pipeline` in Diffusers.
124
125 The checkpoint includes its recommended sampling schedule, so it is ready to use without manually configuring the scheduler. Generation uses CFG=1 by default, and prefix KV caching reuses the text and reference-image context across denoising steps.
126
127 For more details, see the [GitHub repo](https://github.com/QwenLM/Qwen-Image-2.1) and [Qwen-Image-2.1 blog](https://qwen.ai/blog?id=qwen-image-2.1).
128
129 ## Quick Start
130
131 ### Installation
132
133 Install a CUDA-compatible PyTorch build, then install the latest Diffusers source and the following dependencies:
134
135 ```bash
136 pip install git+https://github.com/huggingface/diffusers.git
137 pip install "transformers>=5.17.0" accelerate pillow
138 ```
139
140 This checkpoint requires Diffusers with support for pipeline-configured sampling sigmas, added in [PR #14950](https://github.com/huggingface/diffusers/pull/14950).
141
142 ### Text-to-Image
143
144 Load the checkpoint from Hugging Face:
145
146 ```python
147 import torch
148 from diffusers import QwenImage21Pipeline
149
150 pipe = QwenImage21Pipeline.from_pretrained(
151 "Qwen/Qwen-Image-2.1-Turbo",
152 dtype=torch.bfloat16,
153 ).to("cuda")
154
155 prompt = "A vertically oriented educational chemistry study poster fills the frame, designed like a hand-drawn classroom notes page with a clean white background, thick dark navy-blue outer border, and thin inner navy rectangular border lines. The overall aesthetic is colorful, high-contrast, and hand-lettered in a signature Topper style using marker and watercolor effects. At the top left, a small header in black handwriting reads \"Topper’s Notes\", accented by short decorative navy strokes radiating from it. Centered across the top is the main lesson heading in large stylized black-and-yellow bold lettering: \"Science – Chapter 1:\" followed below by the much larger title \"Chemical Reactions and Equations\". The title has energetic yellow highlighter strokes underneath and small sparkle doodles nearby. At the top right, inside a small outlined page-number box, the label reads \"Page X\". Below the header, a wide rounded rectangle framed in dark navy contains a pink brush-stroke highlight banner with the subheading \"IMPORTANT CHEMICAL EQUATION:\" in black. Inside the same equation panel, the reaction is written very large as \"Fe + CuSO₄ → FeSO₄ + Cu\". The symbol \"Fe\" is highlighted in bright orange with an underline, \"CuSO₄\" is vivid blue and underlined, \"FeSO₄\" is bright green and underlined, and \"Cu\" is metallic bronze and underlined. Beneath the left side of the equation is a lavender rounded label connected by a leader line to the iron component; it says \"Reactants\" and below it \"(before the reaction)\", with \"Reactants\" colored purple and underlined. Beneath the right side is a matching pale-green rounded label connected by a leader line to the products; it says \"Products\" and below it \"(after the reaction)\", with \"Products\" colored green and underlined. Along the lower part of this equation box are stylized ball-and-stick molecule models and atomic orbital drawings: a silver-gray sphere labeled \"Fe\" sits left of center, another cluster shows gray, red, and white spheres representing sulfate groups around blue and green atoms, and a copper-colored sphere appears near the right. Small atom-and-orbit icons sit along the bottom edge of the panel in blue, green, and brown outline styles. The middle portion is enclosed by a large dashed green rounded rectangle titled \"REACTIVITY CONCEPT\" in green uppercase letters with decorative rays. On the left side of this section, a stepped vertical reactivity-series ladder is drawn with alternating colored horizontal blocks. From top to bottom it lists \"K\", \"Na\", \"Ca\", \"Mg\", \"Al\", \"Zn\", \"Fe\" with a green oval callout reading \"(More reactive)\", then continues downward with \"Cu\", \"Ag\", \"Hg\", \"Pt\", \"Au\" with an orange oval callout reading \"(Less reactive)\". Each step is marked by a dark upward-pointing triangular arrow, showing decreasing reactivity downward. In the central-right area, full body text as follows: \"Iron (Fe) is higher in the reactivity series than Copper (Cu). This makes Iron more reactive.\" Below this definition is a visual arrow sequence of overlapping circles: orange \"Fe\" overlaps blue \"Cu\", followed by a red arrow pointing toward the phrase \"→ More displacement ↓ less reactivity\". Under it, a prominent statement in dark handwriting reads \"More reactive element displaces less reactive element\", emphasized with a bright pink marker underline. On the right, a shield-like badge with a navy outline and light interior contains the bold contrasting font label \"DISPLACEMENT REACTION\". Below the badge, full body text as follows: \"A more reactive element displaces a less reactive element from its compound.\" The badge includes an internal miniature illustration of circular electron shells and colored particles, reinforcing the chemistry theme. Directly below the reactivity concept box, a smaller rounded rectangular table titled \"PROPERTY COMPARISON\" compares Iron and Copper. The table has three columns: \"IRON (Fe)\", \"COPPER (Cu)\", and a central property column. Row one reads \"Reactivity series position\" with Iron shown as \"Higher (more reactive)\" and Copper shown as \"Lower (less reactive)\". Row two reads \"Displaces hydrogen from acids\" with Iron marked \"Yes\" and Copper marked \"No\". Row three reads \"Displaced by other metals\" with Iron shown as \"Displaced by Zn, Mg, Al, etc.\" and Copper shown as \"Displaced by Fe, Zn, Mg, etc.\". Row four reads \"Oxidation tendency\" with Iron shown as \"Gets oxidized easily\" and Copper shown as \"Does not get oxidized easily\". Row five reads \"Useful property\" with Iron described as \"Used where strength & corrosion resistance needed\" and Copper described as \"Used where conductivity & corrosion resistance needed\". The table uses thin navy grid lines, orange tinting for the iron column, pale blue tinting for the copper column, black condensed handwriting, and green checkmarks next to positive traits and red crosses next to limitations. The lower half introduces a large illustrated experiment section with a cream highlight behind the title \"BEFORE/AFTER IRON NAIL EXPERIMENT\" in bold dark navy uppercase letters. The section is split visually into a left \"Before\" state and a right \"After\" state. On the left, a beaker filled with bright blue solution contains a shiny silver iron nail standing diagonally. The beaker label reads \"CuSO₄(aq)\" in blue, and the bottom label near the nail reads \"Fe(s)\". A rounded annotation box beside the beaker says \"Initial State: Blue Solution & Silver Nail\". Two glass test tubes stand to the right of this beaker: the upper test tube is filled with blue solution and labeled \"CuSO₄(aq)\", while the lower one contains a pale green liquid. Red and blue directional arrows connect these visual elements to explanations: one red arrow points toward the nail with the text \"Copper gets deposited\", and one blue arrow points toward the liquid with the text \"Solution turns pale green\". On the right side of the experiment section, another beaker shows the after state with the iron nail now coated in brown copper deposits. Its solution is faded to a pale greenish-blue, and the beaker label reads \"FeSO₄(aq)\" in green. A nearby rounded annotation box says \"Final State: Pale Green Solution & Brown Coated Nail\". To the far right are additional test tubes and ion symbols: a test tube containing blue liquid labeled \"CuSO₄(aq)\" with the note \"Blue colour due to Cu²⁺ ions\", and a lower test tube containing pale green liquid labeled \"FeSO₄(aq)\" with the note \"Pale green due to Fe²⁺ ions\". Beside them are ion diagrams labeled \"Cu²⁺(aq)\" and \"Fe²⁺(aq)\", shown as colored charged particles encircled by simple electron-shell rings; the copper ion uses blue particles and the iron ion uses green particles. Small sparkles, arrows, and dashed guide lines emphasize the transformation from blue copper sulfate solution to pale green iron sulfate solution. Near the bottom, a horizontal summary strip has a blue star icon followed by the heading \"SUMMARY\". Full body text as follows: \"In this reaction, Iron (Fe) displaces Copper (Cu) from Copper Sulphate (CuSO₄) solution to form Iron(II) Sulphate (FeSO₄) solution and Copper (Cu).\" To the lower left, a bordered key-points box titled \"KEY POINTS\" lists three starred bullet statements: \"Iron is above copper in the reactivity series.\", \"Fe displaces Cu from CuSO₄ solution.\", and \"Result: Blue solution turns pale green; copper metal deposits on iron nail.\" At the very bottom, inside the footer border, spaced text reads \"CHEMISTRY • CHANGE IS EVERYWHERE • TOPPER’S NOTES\", flanked by small leaf-like ornaments and thin horizontal divider lines. The entire page uses dark navy borders and headings, bright highlighter strokes, watercolor shading, realistic glassware illustrations, and playful scientific doodles to present a chemistry lesson clearly and attractively."
156
157 image = pipe(
158 prompt=prompt,
159 width=1680,
160 height=2512,
161 use_kv_cache=True,
162 generator=torch.Generator("cpu").manual_seed(42),
163 ).images[0]
164
165 image.save("t2i.png")
166 ```
167
168 The pipeline automatically uses the checkpoint's saved **8-step schedule**. The text-to-image example uses **1680 × 2512** portrait resolution; the editing example uses **2048 × 2048**.
169
170 ### Image Editing
171
172 Reuse the pipeline above to turn a yacht sketch into a photorealistic image:
173
174 ```python
175 from diffusers.utils import load_image
176
177 input_image = load_image(
178 'https://huggingface.co/Qwen/Qwen-Image-2.1-Turbo/resolve/main/assets/turbo-boat-input.png'
179 ).convert("RGBA")
180
181 edit_prompt = "Create a photorealistic photograph of the motor yacht from the input sketch, sailing on open sea, captured in full side profile with the camera at water level. Build the vessel with the structure shown in the input sketch: a long low hull; one tall mast stepped forward with a small cross-tree platform near its top; a wheelhouse superstructure with an open bridge wing just behind the mast; a midships deckhouse carrying a row of four rectangular windows with a lifeboat stowed on its roof; a rectangular funnel behind that deckhouse; an aft deckhouse with a row of three rectangular windows and a second lifeboat above it; thin railings along every deck edge; the hull side pierced by one small square window at the bow, an upper line of five round portholes on the raised forward section, and a lower line of nine round portholes along the main hull. Paint the hull black, the superstructures and the mast white, and the funnel yellow. Set the yacht broadside on a calm deep-blue sea with a low gentle swell and a thin white wake trailing from the bow, the straight horizon crossing behind the funnel top, under a clear sky with soft scattered clouds; sunlight from the front left lights the white superstructures and leaves a bright reflection path on the water beside the hull with a soft contact shadow along the waterline; the entire vessel fits inside the frame with open water in the foreground and sky above."
182
183 image = pipe(
184 prompt=edit_prompt,
185 image=input_image,
186 width=2048,
187 height=2048,
188 use_kv_cache=True,
189 generator=torch.Generator("cpu").manual_seed(20261009),
190 ).images[0]
191
192 image.save("edit.png")
193 ```
194
195 ### Sampling
196
197 The recommended 8-step sampling schedule is saved with the checkpoint and loaded automatically. Setting `num_inference_steps` alone does not override it. For experiments, an explicit call-time `sigmas` argument overrides the saved schedule; other schedules have not been evaluated for this checkpoint.
198
199 ### Supported Aspect Ratios
200
201 Use the same resolution presets as Qwen-Image-2.1:
202
203 ```python
204 aspect_ratios = {
205 "1:1": (2048, 2048),
206 "4:3": (2400, 1792),
207 "3:4": (1792, 2400),
208 "3:2": (2528, 1696),
209 "2:3": (1696, 2528),
210 "16:9": (2752, 1536),
211 "9:16": (1536, 2752),
212 }
213 ```
214
215 ## Showcase
216
217 Examples from the 8-step outputs in our Turbo showcase. Each category features one example.
218
219 ### Portrait Photography
220
221 <p align="center">
222 <img src="https://huggingface.co/Qwen/Qwen-Image-2.1-Turbo/resolve/main/assets/turbo-portrait.png" width="48%"/>
223 </p>
224
225 ### Human Poses and Motion
226
227 <p align="center">
228 <img src="https://huggingface.co/Qwen/Qwen-Image-2.1-Turbo/resolve/main/assets/turbo-ballet.png" width="48%"/>
229 </p>
230
231 ### Transparent Image Generation
232
233 <p align="center">
234 <img src="https://huggingface.co/Qwen/Qwen-Image-2.1-Turbo/resolve/main/assets/turbo-rgba.png" width="70%"/>
235 </p>
236
237 ### Typography and Poster Design
238
239 <p align="center">
240 <img src="https://huggingface.co/Qwen/Qwen-Image-2.1-Turbo/resolve/main/assets/turbo-poster.png" width="48%"/>
241 </p>
242
243 ### UI and Information Layout
244
245 <p align="center">
246 <img src="https://huggingface.co/Qwen/Qwen-Image-2.1-Turbo/resolve/main/assets/turbo-ui.png" width="70%"/>
247 </p>
248
249 ### Single-Image Transformation
250
251 **Input**
252
253 <p align="center">
254 <img src="https://huggingface.co/Qwen/Qwen-Image-2.1-Turbo/resolve/main/assets/turbo-figure-input-1.png" width="30%"/>
255 </p>
256
257 **Output**
258
259 <p align="center">
260 <img src="https://huggingface.co/Qwen/Qwen-Image-2.1-Turbo/resolve/main/assets/turbo-figure.png" width="48%"/>
261 </p>
262
263 ### Multi-Reference Composition
264
265 **Input**
266
267 <p align="center">
268 <img src="https://huggingface.co/Qwen/Qwen-Image-2.1-Turbo/resolve/main/assets/turbo-composition-input-grid.png" width="85%"/>
269 </p>
270
271 **Output**
272
273 <p align="center">
274 <img src="https://huggingface.co/Qwen/Qwen-Image-2.1-Turbo/resolve/main/assets/turbo-composition.png" width="70%"/>
275 </p>
276
277 ### Four-Image Interior Composition
278
279 **Input**
280
281 <p align="center">
282 <img src="https://huggingface.co/Qwen/Qwen-Image-2.1-Turbo/resolve/main/assets/turbo-interior-input-grid.png" width="85%"/>
283 </p>
284
285 **Output**
286
287 <p align="center">
288 <img src="https://huggingface.co/Qwen/Qwen-Image-2.1-Turbo/resolve/main/assets/turbo-interior.png" width="70%"/>
289 </p>
290
291 ## License
292
293 This model is licensed under the [Qwen Research License Agreement](https://huggingface.co/Qwen/Qwen-Image-2.1-Turbo/blob/main/LICENSE).
294
295 ## Feedback
296
297 Having issues with Qwen-Image-2.1-Turbo?
298 Our official feedback form connects you directly with the Qwen Image research team.
299 Share your prompts, images, or workflows to help us investigate and improve.
300
301 [**Submit Feedback →**](https://alidocs.dingtalk.com/notable/share/form/v01WgZOZA5DaVQPeqLX_dv19yqvsgs3oebp3pcjys_1qX0QQ0?source=link)
302