Zhipu AI, a Chinese language synthetic intelligence startup, unveiled a brand new AI agent known as AutoLM Rumination on Monday, Reuters reported.
The device can carry out deep analysis duties, generate detailed studies, and assist customers plan varied actions and duties—all at no cost.
The AI agent is powered by the corporate’s proprietary fashions, that are claimed to be nearly as good as DeepSeek R1.
The report additionally famous that the corporate claims its mannequin operates eight occasions sooner whereas utilizing solely one-thirtieth of the assets.
Not too long ago, the startup secured $69 million in Sequence D funding, simply weeks following a $137 million funding spherical. Zhipu is presently valued at $2.74 billion, and it has additionally been beforehand backed by big firms Tencent and Alibaba. The startup, on a number of events, has been backed by Chinese language state-owned entities.
In January, the US commerce division positioned Zhipu and its subsidiaries on the export management entity checklist, stopping them from acquiring US-made parts.
The announcement comes simply weeks after Manus AI, an AI agent from a Chinese language firm, stunned many within the AI ecosystem. This “general-purpose AI agent” outperformed OpenAI’s deep analysis device on the GAIA benchmark, which exams AI brokers on fixing real-world issues.
Manus AI is now out there for a month-to-month price of $39, together with an upgraded plan that prices $199 per thirty days.
The AI ecosystem in China is presently experiencing unprecedented development. Giants like Alibaba and Tencent are unveiling new AI fashions at a rare tempo. Not too long ago, Alibaba launched QwQ 32B, an AI mannequin with 32 billion parameters. Nonetheless, it’s reported to realize efficiency akin to DeepSeek-R1, which has 671 billion parameters (with 37 billion activated parameters).
The mannequin additionally surpasses OpenAI’s o1-mini in a number of benchmarks associated to code, mathematical reasoning, and normal problem-solving duties.
Whereas DeepSeek brought on fairly a storm within the markets a number of months in the past, the corporate doesn’t plan to cease. The corporate up to date the DeepSeek-V3 mannequin, and the DeepSeek V3-0324 now ranks highest in benchmarks amongst all non-reasoning fashions.
Reuters additionally just lately reported that DeepSeek plans to launch R2 “as early as potential”. The corporate initially meant to launch it in early Might however is now considering an earlier timeline.
The publish Chinese language Startup Launches Free AI Agent Powered by Fashions 8x Quicker than DeepSeek-R1 appeared first on Analytics India Journal.