Deep Principle
The organization that released the Hydrogen-1.0 scientific large model. It focuses on the development of AI models for the field of scientific discovery.
Timeline (last 90 days)
The Deep Principle team released Hydrogen-1.0, an open-weight model post-trained on GLM-5.3 and focused on chemistry, materials science and life sciences.
The above FSR benchmark results and cost estimates are based on Deep Principle's own benchmark setup and are not the result of independent third-party audits.
Deep Principle clarified that the five workflow cases it showcased are curated "model + tool" demonstrations and do not represent the overall success rate of autonomous scientific research; the protocol validation is virtual and does not mean physical experiments were actually performed.
Deep Principle released Hydrogen-1.0, an open-weight science model for chemistry, materials science and life sciences, with 753B parameters.
Deep Principle claims Hydrogen-1.0 outperforms GPT-6-Astra on the FrontierScience-Research benchmark, but the result lacks third-party reproduction.
All Facts (5)
The above FSR benchmark results and cost estimates are based on Deep Principle's own benchmark setup and are not the result of independent third-party audits.
50%UnverifiedDeep Principle clarified that the five workflow cases it showcased are curated "model + tool" demonstrations and do not represent the overall success rate of autonomous scientific research; the protocol validation is virtual and does not mean physical experiments were actually performed.
50%UnverifiedDeep Principle claims Hydrogen-1.0 outperforms GPT-6-Astra on the FrontierScience-Research benchmark, but the result lacks third-party reproduction.
50%UnverifiedThe Deep Principle team released Hydrogen-1.0, an open-weight model post-trained on GLM-5.3 and focused on chemistry, materials science and life sciences.
50%UnverifiedDeep Principle released Hydrogen-1.0, an open-weight science model for chemistry, materials science and life sciences, with 753B parameters.
50%