The artificial intelligence landscape is entering into a whole new era, with the explosion of open-sourced language models like DeepSeek R1.
But there is another company that is hoping to lead the charge of open-source language models, and that company is Sentient, a startup company backed by Peter Thiel's Foundation Fund.
Just this week, the company unveiled Open Deep Search, a groundbreaking open-source AI search framework that has already proven itself as a formidable competitor to industry giants like OpenAI’s GPT-4o and Perplexity.
Outperforms AI giants
Sentient’s ODS has set a new benchmark in the AI ecosystem, achieving an impressive 75.3% accuracy on the Frames benchmark—a widely respected test for evaluating factuality, information retrieval, and reasoning capabilities of AI models.
By comparison, OpenAI’s GPT-4o Search Preview scored 50.5%, while Perplexity Sonar Reasoning Pro peaked at 44.4%.
To ensure the integrity of these results, Sentient implemented stringent measures during testing. Researchers were deliberately isolated from the Frames test sets throughout the evaluation process to prevent bias or inadvertent optimization.
However, Himanshu Tyagi, co-founder of Sentient and professor at the Indian Institute of Science, taunted that there is actually no need for testing of their language model because users can literally adopt their code and verify their accuracy.
Sentient's innovative idea to monetize open-source models
While democratizing AI remains at the heart of Sentient’s philosophy, but the company has also tackled one of the most pressing challenges in the industry: how to monetize open-source models effectively without compromising their accessibility.
Sentient’s solution lies in its innovative fingerprinting technology, which allows developers to protect their intellectual property from being stolen without having to make their language model's close sourced. This approach directly addresses a long-standing dilemma that has hindered widespread adoption of open-source solutions in artificial intelligence.
This unique positioning has already garnered significant interest from developers and users alike. Over 1.8 million individuals have signed up on the waiting list for ODS even before its official launch—a testament to its potential impact on the industry.
Beyond just performance benchmarks, Sentient also conducted one of the largest NFT creation campaigns ever seen in the AI space. This initiative enabled over 650,000 participants to acquire fractional ownership in its AI models—a move that aligns with its mission to empower users through decentralized access to cutting-edge technology.
Sewoong Oh, principal researcher at Sentient and professor at the University of Washington, emphasized the scalability and potential of open-source models to outperform its strongest close-sourced competitors
"With suitable architecture, open-source models can easily surpass closed software giants. The results of these benchmarks validate our mission: To create an open ecosystem that benefits all AI developers and users."
A new era ruled by open-sourced AI
The emergence of Sentient and its ODS framework signals a broader shift within the artificial intelligence industry.
With open-source solutions not only catching up but even dominating the AI market today, this disruption could mark a decisive turning point where accessibility and technical performance take precedence over centralized business models controlled by major corporations like OpenAI or Perplexity.