Qwen-Image-2.1vs.LLaDA-Image

Which one is better?

In our own test both are on par. LLaDA-Image has no arena rating yet. Neither has an API price.

Test images

Same task, the first image of each. The task with the biggest difference is preselected.

Qwen-Image-2.1: Product photography, first image from the fixed test run — Qwen-Image-2.190 out of 100

LLaDA-Image: Product photography, first image from the fixed test run — LLaDA-Image35 out of 100

Result for “Product photography”:Qwen-Image-2.1 ahead

Prompt given to both modelsCreate a square premium product photograph of a brushed titanium wristwatch standing upright on dark green marble. The watch has a cream dial, thin black hands set to 10:10, twelve distinct hour markers, a realistic crown, and a dark brown leather strap. Soft window light from the left, controlled reflections, shallow depth of field, no text, no logo, no extra objects.

What the rating looks at: Materials, reflections, and fine details.

“Ahead”: better in all three rating rounds. “Slightly ahead”: a higher value, but not in every round.

Test protocol of the images

Each test attempt uses the same archived prompt and one request. The first output of that attempt counts. Repeats after technical test failures are disclosed for the affected images. Files are normalized to 1,024 × 1,024 pixels without cropping. Provider resolution, randomness, quality tier and prompt processing can still differ. The images are a practical visual test, not a reconstruction of the arena configuration.

Character and detail

Qwen-Image-2.1: manual, 1:1, requested and received 1024x1024, one request, first output, Qwen/Qwen-Image-2.1 base, 40 steps, true CFG 1.0, prompt enhancement disabled, no negative prompt, no reference images, seed=1591420783, official Space commit 5694809f4c56a4e8c40787b5392cecac36493c5d, generated Nov 1, 2026, 8:49 AM. Source

LLaDA-Image: manual, 2.13.0 PyTorch, Diffusers 0.39.0, Transformers 4.57.6; inclusionAI/LLaDA-Image@e4e2703f410f7ddb6ee8d6b09dac6a8ec5093039; native 1024x1024, 1:1; one request, first output; steps=50, guidance=5.0; seed=6428048197886116697; MPS generator; BF16 text/diffusion components, FP32 denoising latent, receipt-reported CPU FP32 native VAE decode; negative_prompt argument omitted (None default, not the manufacturer's empty-string example); MPS community PR @ 5395ce055da9975b09f71e21b865973d409bb3cd; official pinned text encoder uses eager MoE fallback and SDPA attention; other image-component attention paths were not independently captured by this review; no CUDA numerical equivalence claimed; MPS fallback permitted, actual fallback operations not measured; two submissions: the original request stopped at the 60-GiB free-space safeguard before image or latent (supervisor exit 143); one technical same-request/same-seed retry produced the retained first completed output; no quality reroll, generated Nov 3, 2026, 8:50 AM. Source

Protocol gradually-image-comparison-v1. Prompt SHA 256: 985a75b38642a9f0e5e0c0e0c67f1b7f67780d42bd2168fa24e9a7a7a94b0da6

Qwen-Image-2.1 image SHA 256: b29bb7e207ad464957ee060bbf0655104bcbece703360254e48daa3ec35a62d7

LLaDA-Image image SHA 256: a34d96894f5500c5bce2c16cca14ae09748f3aff4d47d15f8aff0d1be5749211

Infographic

Qwen-Image-2.1: manual, 1:1, requested and received 1024x1024, one request, first output, Qwen/Qwen-Image-2.1 base, 40 steps, true CFG 1.0, prompt enhancement disabled, no negative prompt, no reference images, seed=1283422012, official Space commit 5694809f4c56a4e8c40787b5392cecac36493c5d, prior anonymous attempt refused before inference (ZeroGPU quota), completed through authenticated browser, generated Nov 1, 2026, 8:49 AM. Source

LLaDA-Image: manual, 2.13.0 PyTorch, Diffusers 0.39.0, Transformers 4.57.6; inclusionAI/LLaDA-Image@e4e2703f410f7ddb6ee8d6b09dac6a8ec5093039; native 1024x1024, 1:1; one request, first output; steps=50, guidance=5.0; seed=1354677768900879717; MPS generator; BF16 text/diffusion components, FP32 denoising latent, receipt-reported CPU FP32 native VAE decode; negative_prompt argument omitted (None default, not the manufacturer's empty-string example); MPS community PR @ 5395ce055da9975b09f71e21b865973d409bb3cd; official pinned text encoder uses eager MoE fallback and SDPA attention; other image-component attention paths were not independently captured by this review; no CUDA numerical equivalence claimed; MPS fallback permitted, actual fallback operations not measured, generated Nov 3, 2026, 8:56 AM. Source

Protocol gradually-image-comparison-v1. Prompt SHA 256: ed65cb4e6a51e2336f8b149ad4cbb16e020f07d527ad94362e719d2dcf91d94f

Qwen-Image-2.1 image SHA 256: 316e38ed63ff401f86a7ceee9cc65c74413e0e9c7c899a6e6bf453acc2940dac

LLaDA-Image image SHA 256: cad67ad3f1a475d12bc2152840dc98657f81b5c7318c09a9e9fa40fbb433372f

Product photography

Qwen-Image-2.1: manual, 1:1, requested and received 1024x1024, one request, first output, Qwen/Qwen-Image-2.1 base, 40 steps, true CFG 1.0, prompt enhancement disabled, no negative prompt, no reference images, seed=196359631, official Space commit 5694809f4c56a4e8c40787b5392cecac36493c5d, generated Nov 1, 2026, 8:49 AM. Source

LLaDA-Image: manual, 2.13.0 PyTorch, Diffusers 0.39.0, Transformers 4.57.6; inclusionAI/LLaDA-Image@e4e2703f410f7ddb6ee8d6b09dac6a8ec5093039; native 1024x1024, 1:1; one request, first output; steps=50, guidance=5.0; seed=152405017095880227; MPS generator; BF16 text/diffusion components, FP32 denoising latent, receipt-reported CPU FP32 native VAE decode; negative_prompt argument omitted (None default, not the manufacturer's empty-string example); MPS community PR @ 5395ce055da9975b09f71e21b865973d409bb3cd; official pinned text encoder uses eager MoE fallback and SDPA attention; other image-component attention paths were not independently captured by this review; no CUDA numerical equivalence claimed; MPS fallback permitted, actual fallback operations not measured, generated Nov 3, 2026, 1:02 AM. Source

Protocol gradually-image-comparison-v1. Prompt SHA 256: c67fd71e91304458a2feef5c91efa8677260fab0c0099a29a2c52cb579f5b2f3

Qwen-Image-2.1 image SHA 256: 041f559d4261865b327fcae4f8bb0bd3b176c55a3b1255a042ab3e0fc4f9bc4b

LLaDA-Image image SHA 256: 888e173c25c26ff1dcbfd07aad01ffc6a71cc7735099d3e943e60dfddfe31553

Typography and layout

Qwen-Image-2.1: manual, 1:1, requested and received 1024x1024, one request, first output, Qwen/Qwen-Image-2.1 base, 40 steps, true CFG 1.0, prompt enhancement disabled, no negative prompt, no reference images, seed=158818011, official Space commit 5694809f4c56a4e8c40787b5392cecac36493c5d, generated Nov 1, 2026, 8:49 AM. Source

LLaDA-Image: manual, 2.13.0 PyTorch, Diffusers 0.39.0, Transformers 4.57.6; inclusionAI/LLaDA-Image@e4e2703f410f7ddb6ee8d6b09dac6a8ec5093039; native 1024x1024, 1:1; one request, first output; steps=50, guidance=5.0; seed=8778315039365154420; MPS generator; BF16 text/diffusion components, FP32 denoising latent, receipt-reported CPU FP32 native VAE decode; negative_prompt argument omitted (None default, not the manufacturer's empty-string example); MPS community PR @ 5395ce055da9975b09f71e21b865973d409bb3cd; official pinned text encoder uses eager MoE fallback and SDPA attention; other image-component attention paths were not independently captured by this review; no CUDA numerical equivalence claimed; MPS fallback permitted, actual fallback operations not measured, generated Nov 3, 2026, 12:40 AM. Source

Protocol gradually-image-comparison-v1. Prompt SHA 256: 2e012c3d14ee5dd02203a94f394ca80a5038cdebd6032b25a8b8a221f2255419

Qwen-Image-2.1 image SHA 256: 4670269190f492154c05e3f18c4eb9e3938910ca1aca32503b6a1ceb16d39b38

LLaDA-Image image SHA 256: 256a88541c9821a6242bb434fb520d4d6f0793c934c66acd1fd2e8edd8086b5d

The numbers at a glance

Arena ratingfrom votes by users · higher is better

1,227

Qwen-Image-2.1

Rank 14 of 21

n/a

LLaDA-Image

not in the arena

GraduallyBenchour own rating, 0 to 100 · higher is better

Qwen-Image-2.1

Rank 40 of 108

LLaDA-Image

Rank 63 of 108

Price per imagedollars through the API · lower is better

n/a

Qwen-Image-2.1

no API price

n/a

LLaDA-Image

no API price

Ranks among the 21 image models with an arena rating and the 108 with a GraduallyBench run; prices compared with 28 models with an API price.

Qwen-Image-2.1

Subscription: No Alibaba subscription includes this model. Through other providers from ≈ €9.54 a month (Krea Basic, with limited credits). 5 plans in Germany.
Via API: No API price known. Access: local, Web app.
License: Open weights, Qwen Research License: non-commercial use only; commercial use requires a separate license.

LLaDA-Image

Subscription: n/a No plan in Germany names exactly this model.
Via API: No API price known. Access: local.
License: Open weights.

Subscriptions only count plans that name exactly this model. Subscription prices include VAT, API prices are in dollars.

Arena rating (LMArena) from Nov 6, 2026 · GraduallyBench from Nov 3, 2026 · Subscription database from Sep 29, 2026 (38 days old)

All data side by side

The better value is bold. For ratings only when the gap is clear.

Kennzahl	Qwen-Image-2.1	LLaDA-Image
Ratings
Arena ratingfrom votes by users, higher is better	1,227	n/a
Arena rank	Rank 14 of 21	n/a
Arena votesmore votes, a safer rating	19,378	n/a
GraduallyBenchour own rating, 0 to 100	65	48
GraduallyBench by task
Typography and layoutLegible text, hierarchy, and composition	54	62
Product photographyMaterials, reflections, and fine details	90	35
Character and detailAnatomy, clothing, and spatial consistency	34	63
InfographicNumbers, labels, and visual organization	83	33
API price
Per imagelower is better	n/a	n/a
100 images	n/a	n/a
Profile
Released	Sep 20, 2026	Sep 4, 2026
Weightsopen means: runs locally	openQwen Research License: non-commercial use only; commercial use requires a separate license	open
Access	local, Web app	local
Parameters	not published	6 billion
Largest output	2K	not consistently published
Plans in Germanynaming exactly this model	5	n/a

API prices for one image at 1,024 × 1,024 pixels according to Artificial Analysis.

Sources: Qwen-Image-2.1 model source, status source (verified Sep 22, 2026) · LLaDA-Image model source (verified Nov 1, 2026)

Arena ratings by category

From blind votes on LMArena. “Ahead” means the lead is clear.

No shared arena score in the stored snapshot

The stored benchmark database contains no measurement for Qwen-Image-2.1 and LLaDA-Image from the same arena and category. Older model versions are not used as substitutes.

More comparisons

Compare Qwen-Image-2.1 and LLaDA-Image with the models people compare them with most often.

Compare another matchup

First image modelSecond image model

The most popular comparisons

Show all 18 comparisons