top of page

Moonshot AI's Kimi Chatbot Launches Paid Service with Faster Responses

Moonshot AI's Kimi Chatbot provides a premium service with speedier responses. Users can select from six tiers of "top-up" contracts, ranging from 5.2 yuan to 399 yuan. Kimi Chatbot intends to prioritise processing capacity for paying consumers during busy hours.

According to Chinese media outlet Jiemian News, the chatbot, also known as Yuezhi Anmian in Chinese, recently unveiled these new features.

Moonshot AI, based in Beijing, offers six levels of "top-up" contracts that range from 5.2 yuan (US$0.72) for four days to 399 yuan for a year of "priority use." Screenshots posted on China's social media network Weibo reveal pricing information.

Kimi Chatbot, powered by Moonshot AI's self-developed Kimi large language model (LLM), was released in October of last year and is positioned as China's answer to Microsoft-backed start-up OpenAI's ChatGPT. LLMs are the technology behind training generative AI services like ChatGPT.

In a significant upgrade, Kimi Chatbot claimed in March that it could process up to 2 million Chinese characters in a single prompt, a substantial improvement from its previous version, which could only handle 200,000 characters. However, the increase in user numbers caused technological challenges, resulting in the app and website collapsing for hours on March 21 owing to congestion.

Kimi Chatbot's new top-up service intends to prioritise computer resources for users who have paid for the service during peak hours, when computing power is insufficient, according to the company's payment terms.

Because the creation and operation of AI tools demand significant expenditure, businesses are looking into different business models to generate revenue from customers. OpenAI's ChatGPT, which is not available in China and Hong Kong, costs $20 per month for unlimited prompts and early access to new features.

Baidu's chatbot, Ernie Bot, became the first in China to charge for public use. Since November last year, it has been charging 59.9 yuan per month for a single-month membership for Ernie Bot 4.0, which offers higher-resolution image generation for its text-to-image function. However, Baidu's Ernie Bot 3.5 is still available for free public use.

While the majority of the around 200 Chinese-developed LLMs on the market charge corporate users and developers based on API usage, only a few focus on collecting revenue from consumers.

Last Monday, TikTok owner ByteDance commercialised a range of Doubao LLMs, with the Doubao Pro LLM starting at 0.0008 yuan every 1,000-token prompt. In the context of artificial intelligence, a token is a fundamental unit of data that algorithms process, with 1,000 tokens roughly comparable to 750 English words.

Alibaba Group Holding's Qwen-72b-chat LLM costs 0.02 yuan per 1,000 tokens, but some of its other LLMs, including Qwen1.5-110b-chat and Qwen1.5-32b-chat, are presently free for a limited time. It's worth mentioning that Alibaba owns The South China Morning Post.

Baidu's Ernie 4.0 LLM and start-up ZhiPu AI's GLM-4 LLM both charge 0.12 and 0.1 yuan per 1,000 tokens, respectively.

  • Moonshot AI's Kimi Chatbot introduces premium service with faster responses

  • Users can choose from six tiers of "top-up" plans, ranging from 5.2 yuan to 399 yuan

  • Kimi Chatbot aims to prioritise computing power for paying users during peak hours

Source: SCMP

As technology advances and has a greater impact on our lives than ever before, being informed is the only way to keep up.  Through our product reviews and news articles, we want to be able to aid our readers in doing so. All of our reviews are carefully written, offer unique insights and critiques, and provide trustworthy recommendations. Our news stories are sourced from trustworthy sources, fact-checked by our team, and presented with the help of AI to make them easier to comprehend for our readers. If you notice any errors in our product reviews or news stories, please email us at  Your input will be important in ensuring that our articles are accurate for all of our readers.

bottom of page