Z-Image
Tongyi-MAI's 6B single-stream diffusion transformer with a Qwen3 text encoder. Turbo is distilled to 8 model passes; base is the undistilled model, at 50 steps.
Commercial use OK for size turbo: code Apache-2.0, weights Apache-2.0. Measured on our RTX 4090: 4.50 s per 1024×1024 image at 9 steps, 22.7 GiB VRAM peak. We scored it 0.76 on GenEval. safetensors weights, 30.58 GB.
Commercial use OK for size base: code Apache-2.0, weights Apache-2.0. Measured on our RTX 4090: 47.1 s per 720×1280 image at 50 steps, 22.8 GiB VRAM peak. safetensors weights, 19.11 GB.
License
Commercial use OK
Commercial use OK
- Code
- Apache-2.0checked 10 Nov 2026
- Weights
- Apache-2.0all sizes, checked 10 Nov 2026
The repository has no LICENSE file; the model card's metadata says Apache-2.0.
The repository has no LICENSE file; the model card's metadata says Apache-2.0.
On our RTX 4090, measured
4.50s per image
p90 4.50 s, bf16, batch 1, 1024x1024, 9 steps, guidance 0
- VRAM peak
- 22.7 GiB
- Status
- Runs as is
Environment, 10 Nov 2026
NVIDIA GeForce RTX 4090, driver 580.173.02, CUDA 13.0, torch 2.14.1+cu130, Python 3.12.9, diffusers==0.42.1. FlowMatchEulerDiscreteScheduler, seed 0 on cuda. Baseline 791 MiB. 2 warm-up + 10 timed images, median of 3 runs. Scope: predict(): one image from the prompt: text encoding, every denoising step and the VAE decode, to a PIL image; prompt read beforehand.
Accuracy
0.76GenEvalmeasured
GenEval, 2,212 images, bf16, GenEval af4902f · 10 Nov 2026
Reported0.82 GenEval overall on GenEvalSource: technical report (arXiv 2511.22699v5), table 9
47.1s per image
p90 47.1 s, bf16, batch 1, 720x1280, 50 steps, guidance 4
- VRAM peak
- 22.8 GiB
- Status
- Runs as is
Environment, 11 Nov 2026
NVIDIA GeForce RTX 4090, driver 580.173.02, CUDA 13.0, torch 2.14.1+cu130, Python 3.12.9, diffusers==0.42.1. FlowMatchEulerDiscreteScheduler, seed 0 on cuda. Baseline 790 MiB. 2 warm-up + 10 timed images, median of 3 runs. Scope: predict(): one image from the prompt: text encoding, every denoising step and the VAE decode, to a PIL image; prompt read beforehand.
Accuracy
Not yet scored
Reported0.84 GenEval overall on GenEvalSource: technical report (arXiv 2511.22699v5), table 9
Weights
safetensors
- Size
- 30.58 GB
- sha256
95facd593e2549e8252acb571c653d57f7ddb7f1060d4e81712f152555a88804
safetensors
- Size
- 19.11 GB
- sha256
ecd5df7768856215812af84d5cc785ba9d84c6d8c84e6939c7ccf30a2e0d2425
Images for the shared prompts
Same prompts and seed for every model of this task, at each model's official settings. Failures are marked by hand.
14 of 14 prompts made, 3 marked as failures.

Counting
The prompt
Exactly five green apples in a single row on a wooden table, plain white background, product photo.
A correct image: Exactly five apples, all green, in one row on a wooden surface.

Color binding
The prompt
A blue cube resting on top of a red sphere, with a yellow cone standing beside them, on a grey studio floor.
A correct image: Each color on its own shape: a blue cube on top of a red sphere, and a yellow cone beside them. No swapped or mixed colors.

Spatial relations
The prompt
A cat sitting on the left end of a sofa and a dog sitting on the right end, both facing the camera, living room.
A correct image: A cat on the left and a dog on the right, both on one sofa, both facing the viewer.

Text rendering
The prompt
A wooden sign above a bakery door that reads "FRESH BREAD DAILY" in white painted capital letters.
A correct image: The sign reads FRESH BREAD DAILY, spelled exactly and legible, with no extra letters.

Text rendering, several lines
The prompt
A café chalkboard menu with three lines written in chalk: "Coffee 3", "Tea 2" and "Cake 4".
A correct image: Three legible lines, each spelled exactly with its number: Coffee 3, Tea 2, Cake 4. No other words.

Photorealism, landscape
The prompt
Aerial photograph of a winding river through an autumn forest at sunrise, thin mist lying over the water.
A correct image: Reads as a photo from above: a winding river, autumn-colored trees, mist on the water, with no duplicated or melted trees.

Photorealism, food
Fails: Three slices of pork, not twoThe prompt
Close-up photo of a bowl of ramen with a soft-boiled egg cut in half, sliced green onions and two slices of pork, steam rising, on a dark wooden table.
A correct image: A ramen bowl with a halved soft-boiled egg, green onion and two slices of pork, on a dark wooden table.

Transparency and reflection
The prompt
A glass of water on a windowsill, a rainy city street outside the window, the street seen bent and upside down through the water in the glass.
A correct image: A clear glass of water on a sill, rain outside, and the street visibly refracted through the glass.

Hands
The prompt
Close-up of an adult's two hands kneading bread dough on a floured wooden counter, no face in the frame.
A correct image: Two hands with five fingers each and plausible joints, kneading dough on a floured counter. No face.

Fine detail
Fails: Prints a made-up name on the dialThe prompt
Macro photograph of a plain mechanical wristwatch with no brand name, its hands showing ten past ten, the second hand pointing at six.
A correct image: The hour and minute hands show 10:10 and the second hand points at 6. The dial has no brand or logo.

Long, dense prompt
Fails: One cup instead of two, and four books instead of threeThe prompt
A cozy reading nook: a green armchair with a striped blanket over its arm, a sleeping orange cat curled up on the seat, a floor lamp with a yellow shade to the right of the chair, a small round table to the left holding a white teapot and two cups, a stack of three books on the floor, and rain running down the window behind.
A correct image: All of it: green armchair, striped blanket, orange cat asleep on the seat, floor lamp with a yellow shade on the right, round table on the left with a white teapot and two cups, three books stacked on the floor, rain on the window.

Style
The prompt
A watercolor painting of a lighthouse on a rocky coast under a stormy sky, loose brush strokes, visible paper texture.
A correct image: Looks like a watercolor, with washes and paper texture, not a photo: a lighthouse on rocks under a stormy sky.

Negation
The prompt
An empty park bench in fresh snow, with no people and no birds anywhere.
A correct image: A park bench in snow. No people and no birds anywhere in the image.

Unusual combination
The prompt
A giraffe wearing a long knitted red scarf, standing in a snowy city square at night, street lamps glowing.
A correct image: A giraffe with a red knitted scarf around its neck, in a snowy city square at night, with lit street lamps.
14 of 14 prompts made, 3 marked as failures.

Counting
The prompt
Exactly five green apples in a single row on a wooden table, plain white background, product photo.
A correct image: Exactly five apples, all green, in one row on a wooden surface.

Color binding
The prompt
A blue cube resting on top of a red sphere, with a yellow cone standing beside them, on a grey studio floor.
A correct image: Each color on its own shape: a blue cube on top of a red sphere, and a yellow cone beside them. No swapped or mixed colors.

Spatial relations
The prompt
A cat sitting on the left end of a sofa and a dog sitting on the right end, both facing the camera, living room.
A correct image: A cat on the left and a dog on the right, both on one sofa, both facing the viewer.

Text rendering
The prompt
A wooden sign above a bakery door that reads "FRESH BREAD DAILY" in white painted capital letters.
A correct image: The sign reads FRESH BREAD DAILY, spelled exactly and legible, with no extra letters.

Text rendering, several lines
The prompt
A café chalkboard menu with three lines written in chalk: "Coffee 3", "Tea 2" and "Cake 4".
A correct image: Three legible lines, each spelled exactly with its number: Coffee 3, Tea 2, Cake 4. No other words.

Photorealism, landscape
The prompt
Aerial photograph of a winding river through an autumn forest at sunrise, thin mist lying over the water.
A correct image: Reads as a photo from above: a winding river, autumn-colored trees, mist on the water, with no duplicated or melted trees.

Photorealism, food
The prompt
Close-up photo of a bowl of ramen with a soft-boiled egg cut in half, sliced green onions and two slices of pork, steam rising, on a dark wooden table.
A correct image: A ramen bowl with a halved soft-boiled egg, green onion and two slices of pork, on a dark wooden table.

Transparency and reflection
Fails: No street shows through the water in the glassThe prompt
A glass of water on a windowsill, a rainy city street outside the window, the street seen bent and upside down through the water in the glass.
A correct image: A clear glass of water on a sill, rain outside, and the street visibly refracted through the glass.

Hands
The prompt
Close-up of an adult's two hands kneading bread dough on a floured wooden counter, no face in the frame.
A correct image: Two hands with five fingers each and plausible joints, kneading dough on a floured counter. No face.

Fine detail
Fails: The hands are right, but the dial carries made-up letteringThe prompt
Macro photograph of a plain mechanical wristwatch with no brand name, its hands showing ten past ten, the second hand pointing at six.
A correct image: The hour and minute hands show 10:10 and the second hand points at 6. The dial has no brand or logo.

Long, dense prompt
Fails: Four or five books in the stack instead of threeThe prompt
A cozy reading nook: a green armchair with a striped blanket over its arm, a sleeping orange cat curled up on the seat, a floor lamp with a yellow shade to the right of the chair, a small round table to the left holding a white teapot and two cups, a stack of three books on the floor, and rain running down the window behind.
A correct image: All of it: green armchair, striped blanket, orange cat asleep on the seat, floor lamp with a yellow shade on the right, round table on the left with a white teapot and two cups, three books stacked on the floor, rain on the window.

Style
The prompt
A watercolor painting of a lighthouse on a rocky coast under a stormy sky, loose brush strokes, visible paper texture.
A correct image: Looks like a watercolor, with washes and paper texture, not a photo: a lighthouse on rocks under a stormy sky.

Negation
The prompt
An empty park bench in fresh snow, with no people and no birds anywhere.
A correct image: A park bench in snow. No people and no birds anywhere in the image.

Unusual combination
The prompt
A giraffe wearing a long knitted red scarf, standing in a snowy city square at night, street lamps glowing.
A correct image: A giraffe with a red knitted scarf around its neck, in a snowy city square at night, with lit street lamps.
Use it when
You want a commercial-safe text-to-image model that got text, counts and colors right in our prompts, and can give it almost all of a 24 GB card.
Avoid it when
Anything else needs the GPU at the same time: with every component on the card it only just fits in 24 GB.
Sizes
| Size | Per image | VRAM | Accuracy | Weights | Runs | License |
|---|---|---|---|---|---|---|
| turbo | 4.50 s1024x1024 · 9 steps | 22.7 GiB | 0.76 GenEval, measured0.82 GenEval overall paper | 30.58 GB | As is | Commercial OK |
| base | 47.1 s720x1280 · 50 steps | 22.8 GiB | Not yet scored0.84 GenEval overall paper | 19.11 GB | As is | Commercial OK |
Estimated on other GPUs
Estimate, not measured. Worked out from our RTX 4090 run and NVIDIA's spec sheets. Each kind of GPU work is scaled by that GPU's peak for it: tensor cores at the precision the kernels used, CUDA cores, or memory bandwidth. The CPU's part of each call is kept at what it took on our Intel Core i7-10700K. How we estimate
Where a call's time goes on our RTX 4090: tensor cores, BF16 70%, memory-bound 28%.
GeForce (13)
| GPU | Fits | Per image |
|---|---|---|
| GeForce RTX 5090Blackwell, 32 GB GDDR7 | Yes | ≈ 3.3 s |
| GeForce RTX 5080Blackwell, 16 GB GDDR7 | No, needs 23 GB | Doesn’t fit |
| GeForce RTX 5070 TiBlackwell, 16 GB GDDR7 | No, needs 23 GB | Doesn’t fit |
| GeForce RTX 5070Blackwell, 12 GB GDDR7 | No, needs 23 GB | Doesn’t fit |
| GeForce RTX 5060 Ti 16GBBlackwell, 16 GB GDDR7 | No, needs 23 GB | Doesn’t fit |
| GeForce RTX 5060Blackwell, 8 GB GDDR7 | No, needs 23 GB | Doesn’t fit |
| GeForce RTX 4080Ada Lovelace, 16 GB GDDR6X | No, needs 23 GB | Doesn’t fit |
| GeForce RTX 4070 TiAda Lovelace, 12 GB GDDR6X | No, needs 23 GB | Doesn’t fit |
| GeForce RTX 4070Ada Lovelace, 12 GB GDDR6X | No, needs 23 GB | Doesn’t fit |
| GeForce RTX 3090Ampere, 24 GB GDDR6X | Tight | ≈ 8.7–8.8 s* |
| GeForce RTX 3080Ampere, 10 GB GDDR6X | No, needs 23 GB | Doesn’t fit |
| GeForce RTX 3070 TiAmpere, 8 GB GDDR6X | No, needs 23 GB | Doesn’t fit |
| GeForce RTX 3070Ampere, 8 GB GDDR6 | No, needs 23 GB | Doesn’t fit |
- * It fits only just, so other programs on the GPU could push it out.
Workstation (4)
| GPU | Fits | Per image |
|---|---|---|
| RTX PRO 6000 Blackwell Workstation EditionBlackwell, 96 GB GDDR7 | Yes | ≈ 1.8 s |
| RTX PRO 6000 Blackwell Max-Q Workstation EditionBlackwell, 96 GB GDDR7 | Yes | ≈ 2.0 s |
| RTX 6000 Ada GenerationAda Lovelace, 48 GB GDDR6 | Yes | ≈ 2.8–2.9 s |
| RTX A6000Ampere, 48 GB GDDR6 | Yes | ≈ 5.1–5.2 s |
Data center (6)
| GPU | Fits | Per image |
|---|---|---|
| B200 (HGX, one GPU)Blackwell, 180 GB HBM3E | Yes | ≈ 0.47–0.51 s |
| H200 SXMHopper, 141 GB HBM3e | Yes | ≈ 0.87–0.91 s |
| H100 SXMHopper, 80 GB HBM3 | Yes | ≈ 0.99–1.0 s |
| A100 SXM 80GBAmpere, 80 GB HBM2e | Yes | ≈ 2.4–2.6 s |
| L40SAda Lovelace, 48 GB GDDR6 | Yes | ≈ 3.0 s |
| L4Ada Lovelace, 24 GB GDDR6 | Tight | ≈ 8.6–8.8 s* |
- * It fits only just, so other programs on the GPU could push it out.
Where a call's time goes on our RTX 4090: tensor cores, BF16 69%, memory-bound 31%.
GeForce (13)
| GPU | Fits | Per image |
|---|---|---|
| GeForce RTX 5090Blackwell, 32 GB GDDR7 | Yes | ≈ 34 s |
| GeForce RTX 5080Blackwell, 16 GB GDDR7 | No, needs 23 GB | Doesn’t fit |
| GeForce RTX 5070 TiBlackwell, 16 GB GDDR7 | No, needs 23 GB | Doesn’t fit |
| GeForce RTX 5070Blackwell, 12 GB GDDR7 | No, needs 23 GB | Doesn’t fit |
| GeForce RTX 5060 Ti 16GBBlackwell, 16 GB GDDR7 | No, needs 23 GB | Doesn’t fit |
| GeForce RTX 5060Blackwell, 8 GB GDDR7 | No, needs 23 GB | Doesn’t fit |
| GeForce RTX 4080Ada Lovelace, 16 GB GDDR6X | No, needs 23 GB | Doesn’t fit |
| GeForce RTX 4070 TiAda Lovelace, 12 GB GDDR6X | No, needs 23 GB | Doesn’t fit |
| GeForce RTX 4070Ada Lovelace, 12 GB GDDR6X | No, needs 23 GB | Doesn’t fit |
| GeForce RTX 3090Ampere, 24 GB GDDR6X | Tight | ≈ 91 s* |
| GeForce RTX 3080Ampere, 10 GB GDDR6X | No, needs 23 GB | Doesn’t fit |
| GeForce RTX 3070 TiAmpere, 8 GB GDDR6X | No, needs 23 GB | Doesn’t fit |
| GeForce RTX 3070Ampere, 8 GB GDDR6 | No, needs 23 GB | Doesn’t fit |
- * It fits only just, so other programs on the GPU could push it out.
Workstation (4)
| GPU | Fits | Per image |
|---|---|---|
| RTX PRO 6000 Blackwell Workstation EditionBlackwell, 96 GB GDDR7 | Yes | ≈ 19 s |
| RTX PRO 6000 Blackwell Max-Q Workstation EditionBlackwell, 96 GB GDDR7 | Yes | ≈ 20–21 s |
| RTX 6000 Ada GenerationAda Lovelace, 48 GB GDDR6 | Yes | ≈ 30 s |
| RTX A6000Ampere, 48 GB GDDR6 | Yes | ≈ 54 s |
Data center (6)
| GPU | Fits | Per image |
|---|---|---|
| B200 (HGX, one GPU)Blackwell, 180 GB HBM3E | Yes | ≈ 4.4–4.5 s |
| H200 SXMHopper, 141 GB HBM3e | Yes | ≈ 8.6–8.7 s |
| H100 SXMHopper, 80 GB HBM3 | Yes | ≈ 9.9–10.0 s |
| A100 SXM 80GBAmpere, 80 GB HBM2e | Yes | ≈ 24–25 s |
| L40SAda Lovelace, 48 GB GDDR6 | Yes | ≈ 32 s |
| L4Ada Lovelace, 24 GB GDDR6 | Tight | ≈ 93 s* |
- * It fits only just, so other programs on the GPU could push it out.
Alternatives
Something wrong on this page, such as a license or a number? Email us.