Skip to content
ModelsThatRun

Sana-Sprint

NVIDIA's distillation of Sana to one to four steps: a 1.6B linear diffusion transformer with a Gemma 2 text encoder. Makes 1024x1024 images.

Size

Conditional: code Apache-2.0, weights Apache-2.0. Measured on our RTX 4090: 0.27 s per 1024×1024 image at 2 steps, 12.0 GiB VRAM peak. We scored it 0.75 on GenEval. safetensors weights, 9.03 GB.

License

Conditional

Code
Apache-2.0checked 11 Nov 2026
Weights
Apache-2.0all sizes, checked 11 Nov 2026

Apache-2.0 for Sana-Sprint's own weights. The repository also holds its text encoder, Gemma 2 2B IT, which the model card puts under the Gemma Terms of Use: commercial use is allowed, under the Gemma Prohibited Use Policy.

On our RTX 4090, measured

0.27s per image

p90 0.27 s, bf16, batch 1, 1024x1024, 2 steps, guidance 4.5

VRAM peak
12.0 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. SCMScheduler, seed 0 on cpu. Baseline 801 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.

Reported0.24 s per imageA100, bf16, batch 1, 2 steps, a 1024x1024 imageSource: paper (arXiv 2503.09641v4), table 2

Accuracy

0.75GenEvalmeasured

GenEval, 2,212 images, bf16, GenEval af4902f · 11 Nov 2026

Reported0.77 GenEval overall on GenEvalSource: paper (arXiv 2503.09641v4), table 2, 2 steps

Weights

safetensors

Size
9.03 GB
sha256
effac51c73537487ac1505993f0e7fbd292c310b23aae96f2c3e53131e876f2d
Download 1.6b weights

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, 7 marked as failures.

  1. Sana-Sprint 1.6b's image for the prompt: Exactly five green apples in a single row on a wooden table, plain white background, product photo.

    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.

  2. Sana-Sprint 1.6b's image for the prompt: A blue cube resting on top of a red sphere, with a yellow cone standing beside them, on a grey studio floor. Marked as a failure.

    Color binding

    Fails: The colors are right, but the cube stands beside the sphere and the cone sits on the cube
    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.

  3. Sana-Sprint 1.6b's image for 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.

    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.

  4. Sana-Sprint 1.6b's image for the prompt: A wooden sign above a bakery door that reads "FRESH BREAD DAILY" in white painted capital letters. Marked as a failure.

    Text rendering

    Fails: Garbled letters instead of FRESH BREAD DAILY
    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.

  5. Sana-Sprint 1.6b's image for the prompt: A café chalkboard menu with three lines written in chalk: "Coffee 3", "Tea 2" and "Cake 4". Marked as a failure.

    Text rendering, several lines

    Fails: Scribbles and drawings of cups instead of the three 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.

  6. Sana-Sprint 1.6b's image for the prompt: Aerial photograph of a winding river through an autumn forest at sunrise, thin mist lying over the water.

    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.

  7. Sana-Sprint 1.6b's image for 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. Marked as a failure.

    Photorealism, food

    Fails: Three or four pieces of pork, not two slices
    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.

  8. Sana-Sprint 1.6b's image for 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. Marked as a failure.

    Transparency and reflection

    Fails: The water is full of bubbles, and no street shows through it
    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.

  9. Sana-Sprint 1.6b's image for the prompt: Close-up of an adult's two hands kneading bread dough on a floured wooden counter, no face in the frame.

    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.

  10. Sana-Sprint 1.6b's image for the prompt: Macro photograph of a plain mechanical wristwatch with no brand name, its hands showing ten past ten, the second hand pointing at six. Marked as a failure.

    Fine detail

    Fails: The hands don't show 10:10, and the dial has small garbled lettering
    The 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.

  11. Sana-Sprint 1.6b's image for the 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. Marked as a failure.

    Long, dense prompt

    Fails: The table is on the right, with one cup and no teapot
    The 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.

  12. Sana-Sprint 1.6b's image for the prompt: A watercolor painting of a lighthouse on a rocky coast under a stormy sky, loose brush strokes, visible paper texture.

    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.

  13. Sana-Sprint 1.6b's image for the prompt: An empty park bench in fresh snow, with no people and no birds anywhere.

    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.

  14. Sana-Sprint 1.6b's image for the prompt: A giraffe wearing a long knitted red scarf, standing in a snowy city square at night, street lamps glowing.

    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 1024x1024 images in two steps. It got an exact count, left and right, and an unusual combination right in our prompts.

Avoid it when

You need legible text or exact small details: a sign and a menu came out garbled, and a watch showed the wrong time.

Sizes

SizePer imageVRAMAccuracyWeightsRunsLicense
1.6b0.27 s1024x1024 · 2 steps12.0 GiB0.75 GenEval, measured0.77 GenEval overall paper9.03 GBAs isConditional

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 43%, memory-bound 40%, CPU and kernel launches 12%, CUDA cores, FP32 5%.

GeForce (13)
GPUFitsPer image
GeForce RTX 5090Blackwell, 32 GB GDDR7Yes≈ 0.19–0.20 s
GeForce RTX 5080Blackwell, 16 GB GDDR7Yes≈ 0.33 s
GeForce RTX 5070 TiBlackwell, 16 GB GDDR7Yes≈ 0.38–0.39 s
GeForce RTX 5070Blackwell, 12 GB GDDR7No, needs 12 GBDoesn’t fit
GeForce RTX 5060 Ti 16GBBlackwell, 16 GB GDDR7Yes≈ 0.67–0.71 s
GeForce RTX 5060Blackwell, 8 GB GDDR7No, needs 12 GBDoesn’t fit
GeForce RTX 4080Ada Lovelace, 16 GB GDDR6XYes≈ 0.38–0.40 s
GeForce RTX 4070 TiAda Lovelace, 12 GB GDDR6XNo, needs 12 GBDoesn’t fit
GeForce RTX 4070Ada Lovelace, 12 GB GDDR6XNo, needs 12 GBDoesn’t fit
GeForce RTX 3090Ampere, 24 GB GDDR6XYes≈ 0.43–0.44 s
GeForce RTX 3080Ampere, 10 GB GDDR6XNo, needs 12 GBDoesn’t fit
GeForce RTX 3070 TiAmpere, 8 GB GDDR6XNo, needs 12 GBDoesn’t fit
GeForce RTX 3070Ampere, 8 GB GDDR6No, needs 12 GBDoesn’t fit
Workstation (4)
GPUFitsPer image
RTX PRO 6000 Blackwell Workstation EditionBlackwell, 96 GB GDDR7Yes≈ 0.14–0.15 s
RTX PRO 6000 Blackwell Max-Q Workstation EditionBlackwell, 96 GB GDDR7Yes≈ 0.15–0.16 s
RTX 6000 Ada GenerationAda Lovelace, 48 GB GDDR6Yes≈ 0.21 s
RTX A6000Ampere, 48 GB GDDR6Yes≈ 0.31–0.32 s
Data center (6)
GPUFitsPer image
B200 (HGX, one GPU)Blackwell, 180 GB HBM3EYes≈ 0.069–0.091 s
H200 SXMHopper, 141 GB HBM3eYes≈ 0.090–0.11 s
H100 SXMHopper, 80 GB HBM3Yes≈ 0.100–0.12 s
A100 SXM 80GBAmpere, 80 GB HBM2eYes≈ 0.20 s
L40SAda Lovelace, 48 GB GDDR6Yes≈ 0.22 s
L4Ada Lovelace, 24 GB GDDR6Yes≈ 0.53–0.58 s

From the authors

Where to read their own account of it.

Alternatives

Something wrong on this page, such as a license or a number? Email us.

Every model on this photo
⇆
InputOutput