whatmodel.sh / image / NextStep-1-Large-Edit
NextStep-1-Large-Edit
↗ share
+ contribute
NextStep-1-Large-Edit is an open-weights image model from StepFun, released August 2025. Licence: apache-2.0.
Catalogue identity notes
StepFun NextStep-1 autoregressive image model, editing checkpoint (stepfun-ai/NextStep-1-Large-Edit, pinned bac0d9f, Apache-2.0). Edit-Compass ('Nextstep-V1 Han et al. (2025a)'), UnicBench ('NextStep-1 [34]') and WEAVE ('NextStep-1-Large [66] 15B') evaluate NextStep-1 on image-editing benchmarks; NextStep-1-Large-Edit is the only public NextStep-1 checkpoint fine-tuned for instruction editing (NextStep-1-Large and NextStep-1.1 are text-to-image), so the rows are mapped to it; flagged for the owner in held_for_owner. Comparison-only.
Model information sources
Score within the LMArena text-to-image comparison set only.
$ how it compares
The image models ranked either side of NextStep-1-Large-Edit, out of 197 with a score, then more from StepFun.
| # | model | score | where | ||
|---|---|---|---|---|---|
| 152 |  Stable Cascade | 38±4 | local | ||
| 153 |  Janus-Pro-7B | 38±2 | api $0.0020/image +local | ||
| 154 | NextStep-1-Large-Edit | 37±3 | comparison only | ||
| 155 | Step1X-Edit v1.0 | 37±4 | comparison only | ||
| 156 |  SDXL-Lightning (4-step UNet) | 37±3 | local | ||
| more from StepFun | |||||
| 89 | Step1X-Edit v1.2 | 53±2 | local | ||
| 103 | Step1X-Edit v1.2 Preview | 50±5 | comparison only | ||
| 108 | Step1X-Edit v1.1 | 48±3 | local | ||
$ what it's good at
image to image / editing▁▁▃▃▂▃█▃▃▃▁▁ #76 of 81 * · 79% of its score · 2 tests
Trails OmniGen (v1) here, though they rank lower overall.
| test | result | vs the others in its run | counts |
|---|---|---|---|
| Edit-Compass (EN, overall) | 29.0% ✎ | ▃▃▃·▃▆▂█·▃·▃ beats 4/20 | 50% |
| UnicBench (EN, overall) | 64.1% ✎ | ▂····▂▄▂▄·█▂ beats 3/11 | 29% |
other tests1 not in a task · 17% of its score
| test | result | vs the others in its run | counts |
|---|---|---|---|
| WEAVEBench in-context multi-turn editing (Avg) | 53.4% ✎ | █▄▄·▄▄▄▄█·▄█ beats 5/12 | 17% |
$ where to run it
No verified API or download recorded yet · add one
Weights and documentation: huggingface.co/stepfun-ai ↗
$ how the score was worked out · Edit-Compass & EditReward-Com… 50%, UnicEdit-10M: A Dataset and B… 29%, WEAVE: Unleashing and Benchma… 17%
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
| source | results | share of the score |
|---|---|---|
 Edit-Compass & EditReward-Compass: A Unified Benchmark for Image Editing and Reward Modeling | 1 | 50% |
| UnicEdit-10M: A Dataset and Benchmark Breaking the Scale-Quality Barrier via Unified Verification for Reasoning-Enriched Edits | 1 | 29% |
| WEAVE: Unleashing and Benchmarking the In-context Interleaved Comprehension and Generation (arXiv:2511.11434v1) | 1 | 17% |
| prior · pull toward the middle | 3% | |
| weighted average | = 37.2 |
$ improve this page
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