Sapient Intelligence

Sapient Intelligence is a Chinese AI startup developing a non-autoregressive general-purpose large model aimed at surpassing the theoretical limits of conventional LLMs and reaching "L2" — long-range reasoning capability approaching human-like thinking and self-evolution. Founded by Tsinghua Gen-Z graduate Guan Wang and co-founded by Austin, the company has assembled dual R&D centers in Beijing and the Bay Area, recruiting talent from DeepSeek and Richard Sutton's reinforcement learning team .

The Sapient model distinguishes itself architecturally: it incorporates multi-step computation, memory, and tree-search capabilities, and in testing its Sapient H variant has outperformed optimized Llama architectures on complex constraint-satisfaction, high-dimensional optimization, and mathematical reasoning tasks without requiring chain-of-thought data — achieving over 95% accuracy on expert-level Sudoku through self-learning .

On funding, Sapient closed a seed round of tens of millions of USD led by Vertex Ventures (Temasek-backed) with participation from JAFCO Asia, European and US top-tier VCs, and has since raised additional Seed+ funding at a $200M+ valuation within three months of its prior round; Minerva Capital has served as long-term exclusive financial advisor throughout . The capital is directed primarily toward compute and global talent acquisition .

Related coverage: The startup has been profiled by 暗涌Waves as emblematic of a "non-typical" Chinese AI path — founding team from China, day-one global market targeting, international fund backing, yet competing on foundational architecture rather than application layer . Elsewhere's corpus does not contain more recent developments beyond late 2024 on its technical or commercial progress.