whatmodel.sh / image / Step1X-Edit v1.1

Step1X-Edit v1.1

+ contribute
Update model informationAdd a place to run itAdd a test resultPropose a benchmarkCorrect a benchmark scoreEdit data on GitHub

Step1X-Edit v1.1 is an open-weights image model from StepFun, released July 2025. Its published size is 12.43B parameters. Licence: apache-2.0.

Model information sources

· Weights licence ↗ · Parameter count ↗

Diffusion transformer only: OUR COUNT from the safetensors header of the official step1x-edit-v1p1-official.safetensors (pinned revision; 12,427,800,384 BF16). The HF API gives no total for this multi-checkpoint repo; the multimodal encoder and vae.safetensors are not included. Observed 2026-10-07.

48.2±3.2#87 of 144 image modelsfrom 4 benchmark results across 3 sources

Score within the LMArena text-to-image comparison set only.

$ how it compares

The image models ranked either side of Step1X-Edit v1.1, out of 197 with a score, then more from StepFun.

#modelscorewhereRAMreleased
106Latent Diffusion text2img-large (LDM-KL-8-G, LAION-400M) CompVis48±15local~7.5GB2022-04
107Kandinsky 3 AI Forever48±11local~32GB2023-11
108Step1X-Edit v1.1 StepFun48±3local~51GB2025-07
109FuseDream (BigGAN 512) FuseDream authors48±15comparison only2021-12
110VQ-Diffusion (ITHQ) Microsoft48±15local~4.1GB2022-11
more from StepFun
89Step1X-Edit v1.2 StepFun53±2local~51GB2025-11
103Step1X-Edit v1.2 Preview StepFun50±5comparison only2025-09
154NextStep-1-Large-Edit StepFun37±3comparison only2025-08

see all image models · find the ones that fit your machine

$ what it's good at

image to image / editing▁▁·▃▃▃█▃▃▄▁▁ #50 of 62 · 76% of its score · 4 tests

Beats FLUX.1 Kontext [pro] here, though they rank higher overall.

testresultvs the others in its rundifficultycountswhy
UnicBench (EN, overall)69.2% ✎▂····▂▄▂▄·█▂ beats 5/11easy · avg 73%▇▇▇▇▇▇▇▇▇▇▇▇ 30%at its level
KRIS-Bench overall lab51.6% ✎██·········█ beats 0/2medium · avg 54%▇▇▇▇▇▇▇▇▇ 23%at its level · small run
GEdit-Bench overall (G_O, language split unspecified) lab69.7% ✎█·······█··█ beats 0/2easy · avg 73%▇▇▇▇▇▇▇▇▇ 23%at its level · small run
GEdit-Bench-EN overall (G_O, GPT-4.1 judge) lab69.7% ✎▄·····█·▄▄·▄ beats 3/5—not counted yet: no other model in the same run to compare it with
other tests1 not in a task · 20% of its score
testresultvs the others in its rundifficultycountswhy
WEAVEBench in-context multi-turn editing (Avg)66.9% ✎█▄▄·▄▄▄▄█·▄█ beats 11/12medium · avg 65%▇▇▇▇▇▇▇▇ 20%at its level

$ where to run it

Other ways to run it

  1. Run it as published · needs ~50.6 GB, NVIDIA only · from huggingface.co/stepfun-ai
    git clone -b step1xedit https://github.com/Peyton-Chen/diffusers.git && cd diffusers && pip install -e . && pip install transformers==4.55.0 && python -c "import torch; from diffusers import Step1XEditPipeline; from diffusers.utils import load_image; p=Step1XEditPipeline.from_pretrained('stepfun-ai/Step1X-Edit-v1p1-diffusers', torch_dtype=torch.bfloat16); p.to('cuda'); p(image=load_image('input.png').convert('RGB'), prompt='add a ruby pendant to the neck of the girl', num_inference_steps=28, true_cfg_scale=6.0, generator=torch.Generator().manual_seed(42)).images[0].save('0000.jpg')"

Weights and documentation: huggingface.co/stepfun-ai ↗

$ how the score was worked out · stepfun-ai/Step1X-Edit-v1p2 m… 46%, UnicEdit-10M: A Dataset and B… 30%, WEAVE: Unleashing and Benchma… 20%

Each result implies a score on its own, after adjusting for how hard its test is and who ran it. The score is their weighted average, with a small pull toward the middle for thinly tested models. Each test's weight is its counts column above; here it's added up by who ran the tests. How it works

sourceresultsshare of the scorelicence
stepfun-ai/Step1X-Edit-v1p2 model card · lab-reported2▇▇▇▇▇▇▇▇▇▇▇▇▇▇ 46%Apache-2.0
UnicEdit-10M: A Dataset and Benchmark Breaking the Scale-Quality Barrier via Unified Verification for Reasoning-Enriched Edits1▇▇▇▇▇▇▇▇▇ 30%CC-BY-4.0
WEAVE: Unleashing and Benchmarking the In-context Interleaved Comprehension and Generation (arXiv:2511.11434v1)1▇▇▇▇▇▇ 20%CC-BY-4.0
prior · pull toward the middle▇ 4%
weighted average= 48.2

not counted (1)

stepfun-ai/Step1X-Edit repository README resultsGEdit-Bench-EN overall (G_O, GPT-4.1 judge) [correct]69.7%fewer than three models have this test yet, or this source ran it for this model alone

$ improve this page

Update model informationAdd a place to run itAdd a test resultPropose a benchmarkCorrect a benchmark scoreEdit data on GitHub

Use a short GitHub form or edit the data through a pull request. Every number needs a source; here's which ones we can use.

data updated · sources and licences · json