News All Project News Community News Publications Artificial Intelligence Big Techday July 16th, 2025 Showcase Open Hertzlab at ZKM: AI for virtual city design On July 18th, 2025, our Innovation Hacking team will unveil an innovative prototype at the Open Hertzlab in Karlsruhe as part of our cooperation with the Center for Art and Media... On July 18th, 2025, our Innovation Hacking team will unveil an innovative prototype at the Open Hertzlab in Karlsruhe as part of our cooperation with the Center for Art and Media... July 3rd, 2025 Release of DeepSeek-TNG R1T2 Chimera Today, we are releasing the new DeepSeek-TNG R1T2 Chimera as a Tri-Mind Assembly-of-Experts Large Language Model with three parent models: DeepSeek R1-0528, R1, and V3-0324. Today, we release DeepSeek-TNG R1T2 Chimera.This new Chimera is a Tri-Mind Assembly-of-Experts large language model with three parents, namely DeepSeek's R1-0528, R1 and... July 1st, 2025 TNG AI Insight #1: World Foundational Models Today, we introduce you to World Foundational Models (WFMs), large-scale generative AI systems designed to comprehend real-world dynamics, including physics and spatial... Today, we introduce you to World Foundational Models (WFMs), large-scale generative AI systems designed to comprehend real-world dynamics, including physics and spatial... June 25th, 2025 New research paper on the "Assembly of Experts" for TNG's Chimera model The new paper is now published and explains how we constructed our 671B R1T Chimera daughter model in less than an hour of CPU time from known base models. Our new paper “Assembly of Experts: Linear-time construction of the Chimera LLM variants with emergent and adaptable behaviors” has been published on and .The paper explains how... June 12th, 2025 Article "How Long Prompts Block Other Requests - Optimizing LLM Performance" Serving LLMs for over 50 applications, thereby consuming more than 100 million tokens while generating over ten millions tokens per day, requires us to carefully tune our request... Serving LLMs for over 50 applications, thereby consuming more than 100 million tokens while generating over ten millions tokens per day, requires us to carefully tune our request... June 11th, 2025 Article "AI-assisted Java Migration" Our new article shows you how to modernize outdated legacy code with artificial intelligence and migrate it to a new Java version. Our article “AI-assisted Java Migration” will show you how to modernize outdated legacy code and migrate it to a new Java version using Artificial Intelligence.In our new article... June 5th, 2025 24-hour Follow the Sun Hackathon Our 24-hour Follow the Sun Hackathon brought together 19 colleagues from Australia, Germany, Hungary, and the UK. For a whole day, we worked continuously to create an AI... Our 24-hour Follow the Sun Hackathon brought together 19 colleagues from Australia, Germany, Hungary, and the UK. For a whole day, we worked continuously to create an AI... June 2nd, 2025 An update from our robot G1PO Our robot “G1PO” has successfully been taught to walk with the help of reinforcement learning. Over the past few weeks, our Innovation Hacking Team has successfully trained our Unitree G1 Robot 'G1PO' to walk using Reinforcement Learning techniques. We achieved this... May 29th, 2025 DeepSeek-R1T-Chimera: Happy 1B tokens/day! Our R1T Chimera 671B Open-Weights model, composed of DeepSeek R1 and V3-0324, has now been in use for a month. Our R1T Chimera 671B research prototype, constructed from DeepSeek R1 and V3-0324, has now been in the wild for one month.Usage reached up to 1.35 billion processed tokens per day... May 8th, 2025 Upgrade of our computing infrastructure Eight new AMD MI325X GPUs joined our compute cluster of 24 H100. The new Supermicro server is AI beast of a machine having 2 Terabytes GPU memory. It gives us more capacity and... Eight new AMD MI325X GPUs joined our compute cluster of 24 H100. The new Supermicro server is AI beast of a machine having 2 Terabytes GPU memory. It gives us more capacity and... May 2nd, 2025 Release of DeepSeek-R1T-Chimera DeepSeek-R1T-Chimera, an open-weights model that adds R1's reasoning capabilities to DeepSeek AI V3-0324, has been released. On the weekend, we released DeepSeek-R1T-Chimera, an open weights model adding R1 reasoning to DeepSeek AI V3-0324. In benchmarks, it appears to be as smart as R1 but much faster... April 23rd, 2025 Article "Finetuning olmOCR to be a faithful OCR-Engine" We recently carried out AI-supported fine-tuning to automate internal workflows. We recently created a fine-tune of an Optical Character Recognition (OCR) AI model based on olmOCR to help us automate our internal document processing workflows. In our new... Previous 1 2 3 4 Next Previous news can be found in the archive.
