StepFun
14 models across 2 families · Latest: Step 3.7 Flash (2026-05)
One of China's leading AI 'Six Tigers'.
StepFun's portfolio covers 14 active models across 2 current families, spanning coding, rag, and agents. Open a model detail page to compare provider routes and sourced benchmarks.
Covers 6 workload areas across 14 active tracked models; last verified 2026-06-29.
Use it for
- Teams evaluating coding, rag, and agents across this lab's releases
- Comparing model families before committing to a flagship
- Migration and pricing follow-ups across 14 tracked models
Do not use it for
- Choosing a hosting provider without opening a model page for price ladders
Active models
14
Current models from this lab, excluding deprecated ones
Active families
2
Current model families from this lab
Open catalog
0 open
0 open source / 0 open weights
Lowest output price
$0.300 /1M
Cheapest tracked output across active models, per 1M tokens
Latest dated release
2026-05-29
Step 3.7 Flash
Freshness
2026-06-29
Researched 14d ago
Information
Release cadence
Showing 5 recent dated releases (full timeline below). Latest: Step 3.7 Flash (2026-05-29).
Where this lab wins
- Coding: 2 tracked models with SWE-bench / HumanEval-style scores.
- RAG: 2 tracked models with ruler / needle retrieval benchmarks.
- Agentic: 3 tracked models with BFCL, tau-bench, and SWE-bench tool-use coverage.
- Long-context: 7 tracked models with context-token or InfiniteBench-class signal.
Flagship quality / price signal
Flagship: Step 3.5 Flash (best sourced coding quality-per-dollar in this portfolio).
Coding task grade B · swe-bench-verified score 74.4 · cheapest tracked output $0.300 per 1M tokens
StepFun is a Chinese AI company founded in 2023. One of China's leading AI 'Six Tigers'. StepFun ships 2 model families totaling 14 models, with the most recent release Step 3.7 Flash in 2026-05. Notable families include Step and StepAudio 2.5. Use it as a stable reference for lab background, release coverage, and follow-up model pages as they are added. View official API endpoints, benchmark performance, and coding/agent fit for every StepFun model.
About
StepFun (阶跃星辰) is a Chinese AI startup founded in April 2023 that develops large-scale foundation models for text generation, multimodal understanding, audio processing, and agentic tool use. The company has released Step-1 and Step-2 models, StepAudio models, and open-weight Flash models including Step 3.5 Flash and Step 3.7 Flash.
Featured models
| Model | Released | Context | Input price ($/1M) | Output price ($/1M) | License | Openness |
|---|---|---|---|---|---|---|
| Step 3.7 Flash | 2026-05-29 | 256k | $0.20 | $1.15 | Proprietary | Proprietary |
| StepAudio 2.5 Realtime | 2026-05-24 | - | - | - | Proprietary | Proprietary |
| StepAudio 2.5 ASR | 2026-05-22 | - | - | - | Proprietary | Proprietary |
Model families
Recent releases
- Step 3.7 Flash- 2026-05-29
- StepAudio 2.5 Realtime- 2026-05-24
- StepAudio 2.5 ASR- 2026-05-22
- StepAudio 2.5 TTS- 2026-04-16
- Step 3.5 Flash- 2026-01-29
Top comparisons
FAQ
Who founded StepFun and when?
StepFun was founded in 2023 and is associated with Shanghai, China.
What models has StepFun released?
StepFun ships 14 models across 2 families: Step and StepAudio 2.5.
Is StepFun's technology open source?
StepFun's models are proprietary.
Where is StepFun headquartered?
StepFun is headquartered in Shanghai, China.
What is StepFun known for?
One of China's leading AI 'Six Tigers'. Its most prominent tracked family is Step.
How can I access StepFun's models?
StepFun's models are available via NVIDIA NIM, OpenRouter, and StepFun.
Explore related pages
Last reviewed: 2026-06-29. Data sourced from public lab announcements and provider documentation.