Great insights from Henrik Klagges and Robert Dahlke into the construction of their own LLM: R1T2 Chimera at this year's Big Techday! #tngbtd #ai #tngtech #r1t2chimera

Today we release DeepSeek-TNG R1T2 Chimera. This new Chimera is a Tri-Mind Assembly-of-Experts model with three parents, namely R1-0528, R1 and V3-0324. R1T2 operates at a sweet spot in intelligence vs. output token length. It appears to be... * about 20% faster than R1, and more than twice as fast as R1-0528 * significantly more intelligent than R1 in benchmarks such as GPQA Diamond and AIME-24/25, albeit not quite on R1-0528 level * much more intelligent than our first R1T Chimera, and also think-token consistent, which is a major improvement We perceive it as generally well-behaved and a nice persona to talk to. The weights are on @huggingface under the MIT licence. We are looking forward to your experiments and feedback! Thanks to @deepseek_ai for giving their models to the world, to @chutes_ai and @openrouter for hosting R1T, to @WolframRvnwlf for benchmarking it, to @xlr8harder for beta-testing the new Chimera, and to @natolambert for constructive discussions at @aiDotEngineer.

Running powerful AI models locally isn't just about performance - it's about privacy. We benchmarked 𝗤𝘄𝗲𝗻𝟯 and 𝗚𝗲𝗺𝗺𝗮𝟯 on various consumer GPUs and laptops to see how well you can run your own LLM at home or at work. Check out the performance graph below to see how your hardware stacks up, and visit my blog for the full details and Gemma3 results: boredconsultant.com/2025/06/… Thanks to @tngtech for providing access to their AI hardware pool. #Privacy #AI #Qwen3 #LocalAI #TechBlog #Gemma3

Amazing talk on Vibe vs Vise Coding by David Farago yesterday at the 𝐀𝐈 & 𝐂𝐥𝐨𝐮𝐝 𝐈𝐧𝐧𝐨𝐯𝐚𝐭𝐢𝐨𝐧 𝐌𝐞𝐞𝐭𝐮𝐩! 20 min over time people were still demanding to see more of the optional slides and engaging with questions. Thank you David, thanks to all attendants and thanks to @tngtech for hosting the event. #aicloudinnovation #ai #meetup

Follow the white beach flag 😉

Looking forward to meet you this Thursday at the #aicloudinnovation meetup in Karlsruhe. We just increased our max participant count, so you can RSVP again! meetup.com/ai-cloud-innovati…

Thank you @ccssmnn for having me and @alexanderOpalic as speaker at the Typescript x AI Meetup yesterday. We had a great crowd! Also thx to @tngtech for hosting.

𝐂𝐚𝐧 𝐲𝐨𝐮𝐫 𝐥𝐚𝐩𝐭𝐨𝐩 𝐆𝐏𝐔 𝐫𝐮𝐧 𝐦𝐨𝐝𝐞𝐫𝐧 𝐋𝐋𝐌𝐬 𝐥𝐢𝐤𝐞 𝐋𝐥𝐚𝐦𝐚 𝟑 𝐨𝐫 𝐃𝐞𝐞𝐩𝐒𝐞𝐞𝐤-𝐑𝟏? We benchmarked popular models on consumer hardware: from older laptops to dual Nvidia 5090 setups. Check it out 👉 boredconsultant.com/2025/05/… #AI #LLM #Llama3 #DeepSeek #GPU

7/n How about another USB port? USB Hub? All no.

6/n How about FAT32 instead of NTFS? No.

3/n So you download the Windows Media Creation Tool. And after it took ages to start up, you are finally ready to select the USB drive you already have a Windows boot image on. Aaaand: nothing. Apparently the Tool just randomly devices to not read USB drives. answers.microsoft.com/en-us/…