{"version":1,"type":"story","url":"https://digestai.news/story/deep-principle-releases-hydrogen-1-0-a-753b-open-weight-science-model","json":"https://digestai.news/story/deep-principle-releases-hydrogen-1-0-a-753b-open-weight-science-model.json","markdown":"https://digestai.news/story/deep-principle-releases-hydrogen-1-0-a-753b-open-weight-science-model.md","slug":"deep-principle-releases-hydrogen-1-0-a-753b-open-weight-science-model","headline":"Deep Principle releases Hydrogen-1.0, a 753B open-weight science model","summary":"Deep Principle, a Hangzhou-based AI company, has released Hydrogen-1.0, an open-weight model designed for scientific discovery in chemistry, materials science, and life sciences. The 753B parameter model is built by post-training Z.ai’s GLM-5.3 using domain-specific data, laboratory records, and agent discovery trajectories. It supports a 1M-token context window and is available under the MIT License on Hugging Face and ModelScope.\n\nThe company reports that Hydrogen-1.0 outperforms GPT-6-Astra on the FrontierScience-Research benchmark, scoring 38.3% compared to GPT-6-Astra’s 33.3%. It also achieves 93.0% on GPQA-Diamond, roughly matching Kimi-K3. However, Deep Principle notes that the model trails GPT-6-Astra by 3.33 points on NMR structure elucidation tasks. The release aims to provide a reproducible and inspectable alternative to proprietary models for scientific problem-solving.\n\nDeployment requires significant hardware, with testing conducted on 8 GPUs using SGLang. The model supports NVFP4 for NVIDIA B200/B300 GPUs and MXFP4 for AMD MI350X/MI355X. Deep Principle demonstrated the model’s capabilities in lab workflows, including writing Python for molecular dynamics simulations and planning chemical reactions. While full training datasets are not released, the open weights allow researchers to inspect and adapt the model for their specific scientific needs.","keyPoints":["Deep Principle released Hydrogen-1.0, a 753B open-weight model under MIT License.","Model reports 38.3% on FrontierScience-Research, beating GPT-6-Astra's 33.3%.","Requires 8 GPUs for deployment, supporting NVIDIA B200/B300 and AMD MI350X/MI355X."],"whyItMatters":"This release provides a high-performance, open-weight alternative for scientific research, enabling reproducibility and inspection that proprietary models lack. It lowers the barrier for labs to deploy advanced AI for chemistry and materials science without relying on closed APIs.","category":{"slug":"models","name":"Generative AI & Models","url":"https://digestai.news/category/models"},"entities":{"companies":["Deep Principle","Z.ai","NVIDIA","AMD"],"models":["Hydrogen-1.0","GLM-5.3","GPT-6-Astra","Kimi-K3"],"people":[]},"firstPublishedAt":"2026-11-02T18:41:41Z","updatedAt":"2026-11-02T18:41:41Z","sourceCount":1,"hasPrimarySource":false,"sources":[{"outlet":"MarkTechPost","title":"Deep Principle Releases Hydrogen-1.0: A 753B Open-Weight Model for Scientific Discovery","url":"https://marktechpost.com/2026/11/02/deep-principle-hydrogen-1-0-open-weight-scientific-model","publishedAt":"2026-11-02T18:41:41Z","type":"press","primary":false,"lead":true}],"sourceNotes":null,"discussions":[],"thread":null,"cite":{"text":"Digest AI, \"Deep Principle releases Hydrogen-1.0, a 753B open-weight science model\", 2 November 2026, https://digestai.news/story/deep-principle-releases-hydrogen-1-0-a-753b-open-weight-science-model","publisher":"Digest AI","title":"Deep Principle releases Hydrogen-1.0, a 753B open-weight science model","datePublished":"2026-11-02T18:41:41Z","url":"https://digestai.news/story/deep-principle-releases-hydrogen-1-0-a-753b-open-weight-science-model"},"generatedBy":"Written by Digest AI's editorial model from the linked sources; the sources are the record.","license":"Headlines, digests and key points are written by Digest AI and may be quoted with a link to the story page. Linked articles belong to their publishers. Terms: https://digestai.news/terms#reuse"}