StepFun is a Chinese AI ecosystem operated by Shanghai StepFun Intelligence Co., Ltd. Rather than being only a chatbot, it combines everyday AI assistance with image and video creation, voice features, coding help, document analysis, web research, and desktop automation. Most consumer features are available through StepFun AI on the web and mobile devices, while developers can use the StepFun Open Platform through an OpenAI-compatible API. New users may receive a free trial and additional free credits or task rewards, but registration generally requires a Chinese mobile phone number and paid plans are primarily designed for mainland China.
What is StepFun?
StepFun is a broad artificial intelligence product ecosystem developed by Shanghai StepFun Intelligence Co., Ltd., a China-based company established in 2023. Its main consumer product, StepFun AI, provides a conversational assistant that can answer questions, write and rewrite text, analyze files and media, search for information, help with coding, and generate images or video.
The ecosystem also includes AI Studio for experimentation, StepClaw desktop software for agentic computer workflows, Step Code, subscription products, and the StepFun Open Platform for developers. An agentic workflow means the software can carry out a sequence of actions toward a goal rather than only returning a single answer. The exact availability of these features can change as StepFun updates its models and products.
StepFun is most naturally suited to Chinese-language users. Its interfaces, subscriptions, registration process, and much of its documentation are oriented toward mainland China. Some access may be possible from elsewhere, but international users should not assume that every feature, payment option, model, or support channel will be available in their region.
What can an ordinary user do with StepFun?
StepFun AI is designed for a wide range of everyday tasks. You can use it as a general assistant for questions, writing, research, files, media, and creative work. Examples include:
- Ask questions and learn: request explanations, summaries, comparisons, study help, or assistance understanding a difficult topic.
- Write and edit: draft messages, outlines, reports, scripts, and other text, or ask StepFun to rewrite material in a different tone or format.
- Research information: use web-search features to gather information and organize findings. As with other AI search tools, important facts should be checked against reliable sources.
- Work with files: upload supported documents or other files for summarization, extraction, explanation, and analysis.
- Analyze images and video: provide visual material and ask questions about its contents or use it as part of a broader task.
- Create images and video: describe a scene or concept and use StepFun's generation tools for visual content. Image editing is also supported within the wider model ecosystem.
- Use voice features: interact through voice and access speech recognition, text-to-speech, real-time voice, and related audio capabilities where available.
- Get coding assistance: ask for code, explanations, debugging help, or suggestions for improving a program.
These capabilities are not necessarily offered under one identical quota or subscription. Model access, generation limits, and feature availability may depend on the product, account, region, and current plan.
Is StepFun free?
StepFun provides some free consumer access. New users can receive a free trial with full-feature access for a limited period, and additional free credits or daily check-in and task rewards may be available. The newcomer offer can change, so the current promotion should be checked when registering. A Chinese mobile phone number is normally required for registration.
Free access should be understood as an entry point rather than a guarantee of unlimited use. High-demand models, image and video creation, voice services, file processing, and other resource-intensive functions may have separate quotas or restrictions. StepFun does not provide a single universally applicable free allowance across its entire ecosystem in the supplied information.
Consumer subscriptions are offered under the StepFun AI Membership name, with tiers listed as 尝鲜, 入门, 高级, 进阶, and 专业. The platform also lists separate Step Plan tiers called Flash Mini, Flash Plus, Flash Pro, and Flash Max. The available benefits and prices are presented through StepFun's current Chinese subscription pages and may change. Because verified prices are not provided here, prospective users should consult the official subscription page rather than rely on an old price list.
The developer platform uses a different payment model. Public API models are generally billed according to usage, meaning charges depend on how much a developer sends to and receives from a model. StepFun also offers subscription-based Step Plan options and may provide enterprise or bulk-purchase arrangements.
StepFun's most important features
Chat, reasoning, and research
StepFun's conversational models support ordinary questions as well as more involved reasoning and research tasks. Users can ask for a step-by-step explanation, compare alternatives, organize notes, or turn a large request into a practical plan. The service is especially focused on Chinese-language interaction, and its current model catalog includes general language and reasoning models such as Step 5 Preview, Step 3.7 Flash, and Step 3.5 Flash.
Web-search and real-time information features can help with research tasks, but they do not remove the need to verify important claims. Search results and generated answers can be incomplete, outdated, or misinterpreted, particularly for sensitive legal, financial, medical, or professional decisions.
Files, images, and video
StepFun supports file uploads and file analysis, allowing users to ask questions about supplied documents or other supported content. Its multimodal capabilities mean that it can work with more than text, including images and video in supported experiences. This makes it useful for tasks such as explaining a diagram, summarizing material, reviewing visual content, or combining a document with a written request.
The ecosystem also includes image generation and editing, as well as video generation. StepFun's official registration information describes text, image, and video generation through web and mobile products. A separate NextStep-1 announcement documents image-generation and image-editing research. Generation features can have different quotas and access rules from chat, so users should check the current interface before committing to a workflow.
Voice and audio
StepFun lists voice interaction and a range of audio models. These include speech recognition, real-time voice interaction, text-to-speech, voice generation, and music generation. The current platform highlights StepAudio 3 services, including StepAudio 3 Realtime, StepAudio 3 TTS, StepAudio 3 Gen, StepAudio 3 Music, and StepAudio 3 ASR.
Voice and voice-cloning functions deserve additional privacy care. The consumer privacy information states that voiceprint information may be collected when a user actively records material for voice-cloning features. Users should understand what they are uploading and review the applicable privacy terms before submitting personal recordings.
Memory and temporary conversations
StepFun AI includes memory controls that can allow the service to retain useful preferences or context. It also supports temporary conversations. According to the supplied privacy information, temporary conversations are not stored in normal conversation history and are not used for model training or response optimization under the stated policy.
Memory is convenient, but it also means users should review and manage what the service retains. Avoid entering highly sensitive personal, confidential business, or regulated information unless the relevant privacy and retention terms are acceptable for that use.
StepClaw and desktop use
StepClaw is StepFun's agentic desktop offering for Windows and macOS. It is intended for workflows in which an AI system can help operate a computer or coordinate several actions. This can be useful for users who want assistance beyond a chat window, but desktop agents require care because actions may affect files, applications, accounts, or other data on the computer.
StepFun also provides web, iOS, Android, WeChat mini program, Windows, and macOS access across its ecosystem. Not every product is necessarily available on every platform; the official download and product pages should be used to identify the current option.
How to start using StepFun
- Open the official StepFun consumer website or use the supported mobile or WeChat interface.
- Create an account and complete the required verification. Registration normally requires a Chinese mobile phone number.
- Review the current newcomer offer, free credits, or trial conditions shown for the account.
- Start with a simple request, such as asking for a summary, explanation, rewrite, or file analysis.
- Check the selected model and any visible usage quota before using image, video, voice, or other resource-intensive features.
- Review memory, temporary-conversation, and experience-program settings before uploading sensitive content.
Developers should instead register on the StepFun Open Platform, create an API key, and consult the current model catalog and usage documentation. The platform uses an OpenAI-compatible interface, which means some existing tools can be adapted to use StepFun's service, but developers still need to verify the current model names, limits, pricing, and supported features.
Privacy and data-use considerations
StepFun's consumer service processes information needed to operate an AI assistant, including account and device details, usage records, conversation content, uploaded text, voice, images, video, files, links, and generated outputs. The service provides deletion controls, memory controls, and temporary conversations, but those controls should be checked rather than assumed.
The consumer privacy policy states that de-identified and securely processed conversation data may be used to optimize models when the user experience program is enabled. Users can disable that program. Temporary conversations receive separate treatment under the stated policy and are not used for model training or response optimization. Policy wording and settings can change, so users should read the current Chinese privacy policy before submitting confidential material.
The Open Platform has separate privacy and user-agreement documents. API users should consider that prompts, outputs, account details, payment information, and operational records may be processed as part of providing the developer service.
Strengths and limitations
| Strengths | Limitations |
|---|---|
| Strong Chinese-language interaction and a wide range of text, vision, audio, image, and video capabilities. | Registration, payment, interfaces, and feature access are oriented toward mainland China. |
| One ecosystem covers chat, research, file analysis, coding, media creation, voice, and desktop workflows. | Products, model names, limits, and pricing change quickly, making older guides unreliable. |
| Offers consumer applications as well as an OpenAI-compatible developer platform. | English-language access and international support may be less convenient than with globally distributed alternatives. |
| Includes memory controls, temporary conversations, and an experience-program setting. | Users must review privacy settings carefully, especially before submitting confidential data or voice recordings. |
Who should use StepFun?
StepFun is a good candidate for Chinese-speaking users who want one service for conversational help, research, document work, image and video creation, voice interaction, and coding. It is also worth considering for users who want to explore Chinese AI products, use StepFun's current multimodal model range, or build applications through an OpenAI-compatible API.
It may be less suitable for someone who needs a service with simple global signup, broadly available English-language support, stable international billing, independently verified enterprise privacy commitments, or a mature worldwide distribution network. Users whose work depends on fixed model behavior and predictable limits should also be prepared for frequent changes to StepFun's catalog.
Practical assessment
Choose StepFun if Chinese-language performance, multimodal creation, voice tools, and a broad China-focused AI ecosystem are more important to you than global availability. Its strongest reason to choose it is the combination of chat, research, files, coding, image and video generation, audio, and desktop agents in related products.
The main drawbacks are regional access, Chinese-oriented registration and documentation, changing plans and capabilities, and the need to understand its data-use settings. A globally distributed chatbot or productivity platform may make more sense when easy international access, English-first support, predictable billing, or established enterprise controls are the priority.