EveryDev.ai
Sign inSubscribe
Home
Developers

2,521+ AI companies

  • Radar
  • Trending
AI Tools by Topic
  • AI Coding Assistants
  • Agent Frameworks
  • MCP Servers
  • AI Prompt Tools
  • Vibe Coding Tools
  • AI Design Tools
  • AI Database Tools
  • AI Website Builders
  • AI Testing Tools
  • LLM Evaluations
Follow Us
  • X / Twitter
  • LinkedIn
  • Reddit
  • Discord
  • Threads
  • Bluesky
  • Mastodon
  • YouTube
  • GitHub
  • Instagram
Get Started
  • About
  • Editorial Standards
  • Corrections & Disclosures
  • Community Guidelines
  • Advertise
  • Contact Us
  • Newsletter
  • Submit a Tool
  • Start a Discussion
  • Write A Blog
  • Share A Build
  • Terms of Service
  • Privacy Policy
Explore with AI
  • ChatGPT
  • Gemini
  • Claude
  • Grok
  • Perplexity
Agent Experience
  • llms.txt
Theme
With AI, Everyone is a Dev. EveryDev.ai © 2026
    1. Home
    2. Developers
    3. StepFun

    StepFun

    To develop world-class multimodal foundation models and advance toward Artificial General Intelligence (AGI) through deep hardware and platform integration.

    Visit Website

    At a Glance

    1Tool Listed
    7Products
    44Tool Views
    5Capabilities
    Discussions
    Shanghai, ChinaHeadquarters
    2023Est.
    400Employees
    $1.3BRaised
    Focus Areas
    Conversational Agents
    Image
    LLM Orchestration
    Latest News
    StepFun raises record RMB 5 billion in Series B+ funding.Jan 26, 2026
    Megvii co-founder Yin Qi joins StepFun as Chairman.Jan 26, 2026
    Markets
    • Smartphone Manufacturers
    • Automotive Manufacturers (OEMs)
    • Software Developers
    • Individual Consumers
    • +1 more

    AI Tools by StepFun

    (1)
    View StepFun
    StepFun tool icon

    StepFun

    Multimodal AI Platform and API

    Conversational AIImageLLM Orchestration

    Discussions

    No discussions yet

    Be the first to start a discussion about StepFun

    Latest News

    01/26/2026

    StepFun raises record RMB 5 billion in Series B+ funding.

    caixinglobal.com
    01/26/2026

    Megvii co-founder Yin Qi joins StepFun as Chairman.

    techbuzzchina.substack.com
    07/26/2025

    Geely Auto Group and StepFun showcase joint AI solutions at WAIC.

    businesswire.com
    01/16/2025

    StepFun launches Step Reasoner mini high-performance model.

    baike.baidu.com

    Products & Services

    7
    Step-1 / Step-1V
    2024-03

    A trillion-parameter Mixture of Experts (MoE) large language model and its multimodal (vision) variant.

    Step-2
    2024-07

    A trillion-parameter multimodal foundation model designed for advanced reasoning and complex tasks.

    Step-3
    2025-07

    Cutting-edge multimodal reasoning model with 321B total parameters and 38B active parameters.

    Step Reasoner mini
    2025-01

    Reasoning model focused on logical deduction and problem-solving.

    Market Position

    One of China's 'Six Little Tigers' of AI; distinguishes itself by focusing on 'multimodal + terminal' strategies and early integration with major hardware ecosystems.

    Leadership

    Founders

    JD

    Jiang Daxin (姜大昕)

    Founder and CEO. Former Corporate Vice President at Microsoft, General Manager of WebXT S+D, and Chief Scientist at Microsoft Research Asia (MSRA). PhD from Nanyang Technological University.

    Executive Team

    JD

    Jiang Daxin

    Founder and CEO

    AI veteran with nearly 20 years at Microsoft; expert in deep learning and foundation models.

    YQ

    Yin Qi (印奇)

    Chairman

    Co-founder and former CEO of Megvii Technology; joined StepFun in 2026 to oversee strategic roadmap.

    Board of Directors

    YQ
    Yin Qi
    Chairman of the Board

    Founding Story

    StepFun was founded in April 2023 by Dr. Jiang Daxin, a former Microsoft Vice President, and a team of AI veterans. The company was established to build 'China's OpenAI' with a unique focus on multimodal trillion-parameter models that are designed to be integrated directly into hardware devices like smartphones and automobiles from the start.

    Business Model

    Revenue
    Targeted RMB 1 billion (~$140M) in revenue for 2025.

    Revenue Model

    B2B API usage (token-based), enterprise licensing for hardware integration (automotive and smartphones), and premium subscriptions for consumer apps.

    Pricing Tiers

    Standard API
    $0.57 per 1M input tokens

    For Step-3 model (based on OpenRouter/Platform data).

    Standard API (Output)
    $1.42 per 1M output tokens

    For Step-3 model.

    Free Tier
    $0

    Available for consumer products like YueWen with usage limits.

    Rumored to be planning a Hong Kong IPO in 2026.

    Target Markets

    Industries & Segments
    • Smartphone Manufacturers
    • Automotive Manufacturers (OEMs)
    • Software Developers
    • Individual Consumers
    • Enterprise AI Adopters
    Use Cases
    • Smartphones and Terminal Agents
    • Intelligent Automotive Systems (In-car assistants)
    • Consumer Smart Assistants
    • Enterprise API Solutions
    • AI-powered Social and Character Platforms
    Notable Customers
    • Geely Auto Group
    • Oppo
    • Honor
    • ZTE

    Quick Facts

    Headquarters
    Shanghai, China
    Founded
    2023
    Entity Type
    Limited Liability Company
    Employees
    400
    Total Funding
    ~$1.3 Billion+
    Investors
    Pudong Venture Capital, China Life Private Equity
    Office Locations
    Shanghai
    Shanghai

    Funding History

    Series B+>$717 Million (RMB 5 Billion)
    2026-01-26
    ~$4.8 Billion valuation
    Shanghai Pudong Venture Capital
    China Life Private Equity
    Huaqin Technology
    Shanghai Xuhui Capital
    Series BHundreds of millions of dollars
    2024-12
    Fortera Capital
    Early Rounds (Seed/A)Undisclosed
    2023-2024
    Tencent Holdings
    Qiming Venture Partners
    5Y Capital

    History & Milestones

    2026-01-26

    Completed a record-breaking Series B+ funding round of over RMB 5 billion ($717M).

    2025-01-16

    Launched Step Reasoner mini, a high-performance reasoning model.

    2025-07-25

    Introduced Step-3 foundational model at the World Artificial Intelligence Conference (WAIC).

    2024-03

    Released Step-1 (1 trillion parameter MoE model) and Step-1V (multimodal model).

    2024-07

    Previewed Step-2, a trillion-parameter multimodal model.

    Key Capabilities

    5
    Trillion-parameter Mixture-of-Experts (MoE) architecture
    Native multimodal capabilities (vision-language)
    Contextual speech synthesis (TTS)
    Advanced logical reasoning
    Hardware-level optimization for on-device AI

    Integrations & Partnerships

    Platform Integrations

    • StepFun Open Platform (API)
    • Geely Smart Cockpit
    • Smartphone OS integrations (ColorOS, MagicOS)
    • OpenRouter

    Key Partnerships

    Geely Auto (Strategic Partner)
    Oppo
    Honor

    Connect

    Website
    stepfun.com

    AI Topics

    3

    StepFun focuses on these topics:

    Conversational Agents(1)
    Image(1)
    LLM Orchestration(1)
    Back to all developersSuggest an edit