Get the latest setups
Workflows/Create images locally with Qwen-Image-2.1
Images / WORKFLOW

Create images locally with Qwen-Image-2.1

Create an image with the official Qwen Image 2.1 ComfyUI template, or choose the publisher Turbo pipeline for a faster eight-step route.

Verified sourceSource checked 10/12/2026
LOCALRENTED GPU
Freshness 89 out of 100. Repository age: 31 days old, Sep 14, 2026. Measures age, not quality.
Popularity 74 out of 100. 159,495 downloads in 30 days; 3,315 lifetime likes. Checked Oct 15, 2026. Measures adoption, not quality.
Arena Open Source Image generation: rank 1 of 19. Ranked model: qwen-image-2.1. Original model ranking, not a benchmark of this package. Export dated Oct 7, 2026. Source checked Oct 15, 2026. Human preference ranking, not hardware suitability.
Category rankings & sources

Each position belongs to a separate Arena category. These are human preference rankings—not hardware recommendations. We do not combine the categories into one score.

Image generation#1of 19 open models
qwen-image-2.1alibabaLicense: qwen-research

Original model ranking. This does not benchmark the selected quantization or package.

Export dated Oct 7, 2026Checked: Oct 15, 2026Overall board: #19
View official Image generation board
Image editing#1of 10 open models
qwen-image-2.1alibabaLicense: qwen-research

Original model ranking. This does not benchmark the selected quantization or package.

Export dated Oct 6, 2026Checked: Oct 15, 2026Overall board: #18
View official Image editing board

Rankings reflect the dated export; live Arena boards can change. Open-source positions use Arena’s open-model filter. The overall position also includes the other models on that category board. Availability of weights and license terms still need to be checked separately.

What do these scores mean?

Two ways to compare models. Neither score measures answer quality, benchmark performance, or whether a model fits your machine.

Freshness · release age

A new release starts at 100. The score halves every 180 days: 100 × 2^(−age in days / 180). When the release date is unavailable, a repository date is a fallback—not proof of when the model was released. Missing dates stay unscored.

Check the date source
Popularity · adoption

Downloads over the past 30 days carry 80% of the score; lifetime likes carry 20%. Both use a logarithmic scale, capped at 10 million downloads and 100,000 likes. If likes are unavailable, downloads carry all the weight. Missing download statistics stay unscored.

round(80 × min(1, log10(1 + downloads) / 7) + 20 × min(1, log10(1 + likes) / 5))

Check the statistics source
Scores as of Oct 15, 2026. Statistics checked Oct 15, 2026. Statistics are marked for refresh after 72 hours.Full scoring method

Before you begin

Hardware

These are capacity starting points for the complete pipeline, not speed tests. The ComfyUI Base route and publisher Turbo route use different files and runtimes.

MachineQwen-Image-2.1 · 7B BaseQwen-Image-2.1 Turbo · 7B
NVIDIA GPU · 16 GB VRAMPossible with ComfyUI offload; start at 1024 px and check memory.Check complete BF16 pipeline fit; a small package size alone is insufficient.
NVIDIA GPU · 24 GB VRAMGood starting candidate at 1024 px; verify actual peak.Candidate; verify CUDA memory and leave headroom for encoders.
NVIDIA GPU · 32 GB VRAMGood starting candidate; 2K output costs more memory.Good capacity candidate; benchmark the exact output size.
AMD Radeon · 24 GB VRAMConditional on supported ComfyUI ROCm setup; check all nodes.No documented Turbo ROCm route here; use NVIDIA rental.
Intel Arc · 16 GB VRAMNo validated Qwen 2.1 pipeline in this guide; use NVIDIA rental.No validated Turbo Intel route; use NVIDIA rental.
Apple Silicon · 64 GB unified memoryComfyUI Mac route available; start at 1024 px and check pressure.No validated Turbo MPS route; choose Base or NVIDIA rental.
CPU only · 64 GB RAMNot a practical first-run path for this guide; rent a GPU.No CPU route documented here; rent a GPU.
RunPod · NVIDIA GPU · 48 GB VRAMGenerous memory candidate; check storage and ComfyUI nodes.Generous candidate for BF16; verify runtime before a long rental.

Start small, then raise resolution after a successful render. Source documentation verifies package files and settings; these rows are estimates, not YouRunAI benchmarks.

IMAGE GENERATION · PRIVATE SETUPSOURCE-LINKED SETUP

From prompt to an image you can keep.

Choose the maintained ComfyUI template for Base or the publisher’s faster Turbo checkpoint. The package selector changes every step.

Choose a package Load the visual workflow Check the result
01
BEFORE YOU BEGIN

Choose your package

Select the weights you want. The instructions, downloads and settings change with your choice.

Save your machine in My Hardware to get an automatic starting choice. You can always choose any package yourself.

CURRENT SETUP

Qwen-Image-2.1 · 7B · official

Begin at 1024 px and 25 steps. A 64 GB Apple Silicon Mac is an estimated starting point for the complete ComfyUI pipeline, not a measured speed result. Increasing resolution and image references raises full-pipeline memory.

Freshness 89 out of 100. Repository age: 31 days old, Sep 14, 2026. Measures age, not quality.
Popularity 74 out of 100. 159,495 downloads in 30 days; 3,315 lifetime likes. Checked Oct 15, 2026. Measures adoption, not quality.
Arena Open Source Image generation: rank 1 of 19. Ranked model: qwen-image-2.1. Original model ranking, not a benchmark of this package. Export dated Oct 7, 2026. Source checked Oct 15, 2026. Human preference ranking, not hardware suitability.
Category rankings & sources

Each position belongs to a separate Arena category. These are human preference rankings—not hardware recommendations. We do not combine the categories into one score.

Image generation#1of 19 open models
qwen-image-2.1alibabaLicense: qwen-research

Original model ranking. This does not benchmark the selected quantization or package.

Export dated Oct 7, 2026Checked: Oct 15, 2026Overall board: #19
View official Image generation board
Image editing#1of 10 open models
qwen-image-2.1alibabaLicense: qwen-research

Original model ranking. This does not benchmark the selected quantization or package.

Export dated Oct 6, 2026Checked: Oct 15, 2026Overall board: #18
View official Image editing board

Rankings reflect the dated export; live Arena boards can change. Open-source positions use Arena’s open-model filter. The overall position also includes the other models on that category board. Availability of weights and license terms still need to be checked separately.

What do these scores mean?

Two ways to compare models. Neither score measures answer quality, benchmark performance, or whether a model fits your machine.

Freshness · release age

A new release starts at 100. The score halves every 180 days: 100 × 2^(−age in days / 180). When the release date is unavailable, a repository date is a fallback—not proof of when the model was released. Missing dates stay unscored.

Check the date source
Popularity · adoption

Downloads over the past 30 days carry 80% of the score; lifetime likes carry 20%. Both use a logarithmic scale, capped at 10 million downloads and 100,000 likes. If likes are unavailable, downloads carry all the weight. Missing download statistics stay unscored.

round(80 × min(1, log10(1 + downloads) / 7) + 20 × min(1, log10(1 + likes) / 5))

Check the statistics source
Scores as of Oct 15, 2026. Statistics checked Oct 15, 2026. Statistics are marked for refresh after 72 hours.Full scoring method
View weight source
02
THE WORKFLOW

Set up Qwen-Image-2.1 · 7B on your machine

Choose your computer first, then its GPU path. Follow the visual setup and import the linked workflow.

CHOOSE YOUR SETUPWindows
02

Put the exact model files in place

Download the diffusion checkpoint, base text encoder, and VAE into these ComfyUI/models folders. The prompt-enhancement encoder is optional while refine_prompt is off. Restart ComfyUI after placing files.

Put these files in ComfyUI/models/3 required files
diffusion_models/
qwen_image_2.1_int8_convrot.safetensorsInt8 diffusion checkpoint loaded by the official templateGet file
text_encoders/
qwen3vl_8b_int8_convrot.safetensorsBase text and image encoderGet file
vae/
qwen_image_2.1_vae_bf16.safetensorsImage decoderGet file
04

Describe an image and choose a size

In the Qwen Image 2.1 Text to Image subgraph, select the exact loaded model, enter a scene prompt, start at 1.0 megapixel (about 1024×1024), 25 steps, CFG 1, Euler/simple, and record the seed. Leave prompt enhancement off for the first run.

05

Generate and save your image

Click Queue. Open the generated PNG in ComfyUI/output and compare it to the prompt. Save the seed and size alongside the image so a second run can be reproduced.

03
SUCCESS CHECK · Qwen-Image-2.1 · 7B

A saved, repeatable PNG

Open the PNG in ComfyUI/output or the Turbo script folder. Confirm it follows the prompt and record the package, seed, resolution and settings.

The base and Turbo packages are distinct checkpoints. The ComfyUI template loads a Comfy-Org repackaging of the Qwen publisher weights. Turbo uses Qwen’s own Diffusers example. GPU advice is estimated from full-pipeline components, not a site-run benchmark.

Back to setup steps

When it doesn’t go to plan

ComfyUI shows missing nodes or a missing model.

Update ComfyUI and compare the three required filenames and their folders with the official Qwen guide. Restart after adding files; keep prompt enhancement off until its optional encoder is installed.

The first render is very slow or exhausts VRAM.

Check the actual input pixel count and start near 1024×1024, close other GPU apps, and use the int8 ComfyUI template. Turbo needs its separate BF16 pipeline and sufficient GPU memory.

Turbo fails when loading Diffusers.

Install Diffusers from current source as the Qwen Turbo card requires for saved sampling sigmas; confirm the selected model ID is Qwen/Qwen-Image-2.1-Turbo, not Base.

REFERENCE LIBRARY

Sources and files

Original instructions, model files, and compatibility notes behind this setup.

3 SOURCES
01ComfyUI Qwen Image 2.1 setup, files and templatesdocs.comfy.org02Qwen publisher Base modelhuggingface.co03Qwen publisher Turbo setup and eight-step schedulehuggingface.co

The model behind this workflow

Qwen-Image 2.1

Keep exploring

Text
SETUP GUIDE

Qwen3.5 · 9B private chat setup

Source-backed package and platform-route plan for private local chat with Qwen3.5 · 9B. Preserves the official Ollama 9B package, a smaller 4B family fallback, and a clearly labeled community uncensored GGUF option.

Freshness 41 out of 100. Release age: 230 days old, Feb 27, 2026. Measures age, not quality.
Popularity 93 out of 100. 8,591,266 downloads in 30 days; 2,171 lifetime likes. Checked Oct 15, 2026. Measures adoption, not quality.
SOFTWAREOllama
LOCALRENTED GPU
Estimated
Coding
SETUP GUIDE

Build a private coding assistant with Qwen3.8 27B

Run Qwen3.8 27B with Ollama, connect it to Cline in VS Code, and make one reviewable code change in a tiny local project.

Freshness 76 out of 100. Repository age: 71 days old, Aug 5, 2026. Measures age, not quality.
Popularity 95 out of 100. 7,363,968 downloads in 30 days; 17,749 lifetime likes. Checked Oct 15, 2026. Measures adoption, not quality.
Arena Open Source Web coding: rank 10 of 55. Ranked model: qwen3.8-27b. Original model ranking, not a benchmark of this package. Export dated Oct 13, 2026. Source checked Oct 15, 2026. Human preference ranking, not hardware suitability.+3
SOFTWAREOllama + Visual Studio Code + Cline + Python
LOCALRENTED GPU
Estimated
Images
SETUP GUIDE

Create images locally with FLUX.2 Klein

Generate images from text prompts in ComfyUI with the fast distilled or more adjustable base 4B model. Save your results and reuse the visual workflow.

Freshness 35 out of 100. Release age: 273 days old, Jan 15, 2026. Measures age, not quality.
Popularity 76 out of 100. 423,736 downloads in 30 days; 1,043 lifetime likes. Checked Oct 15, 2026. Measures adoption, not quality.
SOFTWAREComfyUI
LOCALRENTED GPU
Verified source