Sentient, a year-old non-profit backed by Peter Thiel’s Founders Fund, has released Open Deep Search (ODS), an open-source framework that leverages existing LLMs to enhance search and reasoning capabilities. Essentially a system of custom plugins and tools, ODS works with DeepSeek’s open-source R1 model as well as proprietary systems like OpenAI’s GPT-4o and Anthropic’s Claude to deliver advanced search functionality. That modular aspect is in fact ODS’s main innovation, its creators say, claiming it beats Perplexity and OpenAI’s GPT-4o Search Preview on benchmarks for accuracy and transparency.

“For enterprises looking for customizable AI search tools, ODS offers a compelling, high-performance alternative to closed commercial solutions,” VentureBeat posits.

The ODS approach is detailed in a research paper. The backbone relies on general AI architectures, organizing and applying them in novel ways. ODS-v1 utilizes ReAct and Chain-of-Thought reasoning, while ODS-v2 uses Chain-of-Code and CodeAct. Each guides the third-party LLM in generating accurate responses.

“Most innovation in AI search has happened behind closed doors. Open-source efforts have historically lagged in usability and performance,” Sentient co-founder Himanshu Tyagi tells VentureBeat. “ODS aims to close that gap, showing that open systems can compete with, and even surpass, closed counterparts on quality, speed, and flexibility.”

Decrypt calls it “America’s answer to China’s DeepSeek.” Though unlike DeepSeek, ODS isn’t itself an LLM; like the Chinese model it “challenges the ‘dominance of closed AI systems as the U.S. approaches its own ‘DeepSeek moment,’” Decrypt quotes Sentient explaining in a statement. The ODS code is available on GitHub.

It has customization options for go-to sources like Wikipedia, ArXiv and PubMed, and can be set up to prioritize authoritative sources when conflicting information surfaces.

“By providing people and brands full access to the underlying algorithms, Sentient wants search to remain community-driven, rather than controlled by a single entity,” writes Adweek, which also interviewed Tyagi.

The results speak for themselves, according to VentureBeat, which says “both ODS-v1 and ODS-v2, when combined with DeepSeek-R1, outperformed Perplexity’s flagship products. Notably, ODS-v2 paired with DeepSeek-R1 surpassed the GPT-4o Search Preview on the complex FRAMES benchmark and nearly matched it on SimpleQA.”

The approach has landed the non-profit a valuation of $1.2 billion, per Adweek.