— Image Tool

StepFun

Last updated 2026-07-15 · Reviewed by ToolForge Editorial

Chinese AI lab behind Step-1 image, Step-2 video, and Step-1o audio - fast-rising open-weight contenders.

★ 4.2/5 · Tens of millions · Since 2023 · Free tier

Free chatbot

Try StepFun → Read full review

A new frontier lab from China

StepFun is a Shanghai AI startup whose Step series rivals the best image and video generators. Step-1X produces editorial-grade images, Step-2 is a strong text-to-video model, and Step-1o handles real-time voice. They release open weights and a free consumer app, making frontier Chinese generative AI accessible worldwide.

Who it's for: Creators and developers who want high-quality Chinese-style generation with open-weight flexibility.

Key features

🖼️ Step-1X images

Photoreal and illustrated output competitive with Midjourney.

🎬 Step-2 video

Text-to-video with strong motion coherence.

🎙️ Step-1o audio

Low-latency conversational voice.

🔓 Open weights

Key models released for self-hosting.

The honest take

✓ What works

High-quality image and video
Free consumer app
Open-weight releases
Fast iteration

✗ What doesn't

English prompts sometimes weaker
Less known outside China
Video still behind Kling
Ecosystem maturing

Verdict

StepFun is one of the most exciting open-weight generative labs of 2026. Worth a look if you want Midjourney-class quality with self-host freedom.

💡 Transparency: This review contains affiliate links. If you sign up through our link, we may earn a commission at no cost to you. We only recommend tools we use ourselves. Full disclosure.