AI News HubLIVE
站内改写

5 Billion Tokens Free! World's First Commercial AI Host Launched, Unleashing Token Consumption

Lenovo launches the world's first commercial AI host series, designed for one-person companies (OPC) and growing enterprises. By combining local and cloud hybrid architecture, it addresses high token costs and data security issues, offering generous token bonuses and out-of-box experience.

Article intelligence

EngineersAdvanced

Key points

  • Lenovo unveils three AI hosts: mini 100, 300, and Pro 700, catering from individuals to teams.
  • Local inference plus cloud elasticity reduces token costs by 70%-95%.
  • Purchase includes 200 million to 500 million free tokens, removing usage limits.
  • The 'Star River Plan' aims to build an ecosystem serving over 5 million OPX (One Person X).

Why it matters

This matters because lenovo unveils three AI hosts: mini 100, 300, and Pro 700, catering from individuals to teams.

Technical impact

May affect model selection, inference cost, product capability, and evaluation benchmarks.

On May 29, 2026, during a live product launch event, Lenovo released the world's first commercial AI host series: the Lenovo Baishih AI Host. The series includes three models—mini 100, 300, and Pro 700—specifically designed for One Person Companies (OPC) and growing enterprises. These devices aim to solve the pain points of high token costs, data security concerns, and complex deployment in the AI era.

As AI agents become more deeply integrated into business workflows, token consumption skyrockets. Traditional cloud-based pricing models lead to linearly growing costs, deterring full automation. Moreover, agents handling core business data face severe security risks. Local deployment, while cheap, requires specialized AI engineers and time-consuming setup. Lenovo's AI host offers a balanced middle ground: it acts as an intelligent hub, automatically routing requests between local and cloud models based on task complexity. This hybrid approach guarantees stable computing power with a one-time hardware investment, after which only electricity costs remain.

The mini 100, with a 0.5-liter compact body and low noise, is ideal for individuals. It reduces token costs for tasks like market analysis by 70%-95% thanks to context compression. The 300 model features a local 35B multimodal model and edge-cloud scheduling, achieving over 80% cost savings for project management tasks. The top-tier Pro 700 can deploy up to 122B multimodal models locally, with 8x concurrent inference and 99.9% availability. It also supports multi-device clustering and unified management of various AI tools used by employees.

Lenovo is not just selling hardware; it is building an ecosystem. The launch event introduced the 'Star River Plan' with over 100 million global exposures, up to 50 million yuan in investment support, and massive token incentives for partners. The goal is to host 10,000+ service delivery partners, 50,000+ digital roles, and 100,000+ Skills, ultimately serving over 5 million OPX entities (One Person X, where X can be company, department, or individual).

In essence, Lenovo's AI host is likened to Gutenberg's printing press for the AI age: it democratizes access to AI computing, enabling creators and coders to turn ideas into reality without being constrained by token budgets or technical barriers.