Overview
A 14B autoregressive image generation and editing model using continuous image tokens, exploring image modeling without discrete VQ tokens.
NextStep-1
A 14B autoregressive image generation and editing model using continuous image tokens, exploring image modeling without discrete VQ tokens.
MODEL NUMBER352
- First public date
- 2025-08-14
- Organization
- StepFun
- Family
- Step Image
01
Specifications
Only published specifications are shown.
- Total parameters
- 14B
- Architecture
- Dense
- Input modalities
- Text · Image
- Output modalities
- Image
- Open weights
- Yes
- License
- Apache-2.0
- Status
- Legacy
02
Overview
Why it mattered
NextStep-1 extends StepFun’s image line from editing into unified autoregressive generation and editing.
03
Architecture & capabilities
Model capabilities
Image generationImage editing
04
Release history
The NextStep-1-Large series released inference code and image-model weights.
View source05
Related models
Select a card to view the related model.
06
Sources
Suggest a correction
Copy this template, add the field, proposed value, and primary source, and share it with the archive maintainer.