--- format: "aidr-story-markdown/v1" id: "a20cd4f411c207db72d94db18f83047b8e1b4d05e4c7be49dfc646c7149bb427" canonical_url: "https://aidr.today/a20cd4f4?lang=en" title: "Together's Post-Training B200 GPU Rates Drop to $9 an Hour for AI Fine-Tuning" lang: "en" requested_lang: "en" available_langs: ["en","vi"] translation_fallback: null fallback_fields: [] published_at: "2026-10-29T18:54:15.000Z" category: "Tools" topics: ["llm","devtools"] source_urls: ["https://huggingnews.com/ai/update-togethers-post-training-b200-gpu-rates-drop-to-9-an-hour-for-ai-f-8e656a44","https://marketbrief.now/ai/update-togethers-post-training-b200-gpu-rates-drop-to-9-an-hour-for-ai-f-8e656a44"] summary: "Together reduced its Post-Training platform rate for Nvidia B200 GPU hours to $9, down from $11, for the first eight weeks of use. The platform added the MiniMax M3.1 Flash model and supports supervised fine-tuning and reinforcement learning to convert open-weight models into proprietary tools. Deel's data science team used the service to fine-tune its Pangaea K1 model with over 43,500 pages of knowledge, while Jack & Jill built a candidate prediction model matching closed-source AI quality." --- # Together's Post\-Training B200 GPU Rates Drop to $9 an Hour for AI Fine\-Tuning > [Open the canonical story]() **Published:** 2026-10-29T18:54:15.000Z **Category:** Tools **Topics:** llm, devtools ## Summary Together reduced its Post\-Training platform rate for Nvidia B200 GPU hours to $9, down from $11, for the first eight weeks of use\. The platform added the MiniMax M3\.1 Flash model and supports supervised fine\-tuning and reinforcement learning to convert open\-weight models into proprietary tools\. Deel's data science team used the service to fine\-tune its Pangaea K1 model with over 43,500 pages of knowledge, while Jack & Jill built a candidate prediction model matching closed\-source AI quality\. ## Sources - [Story source]() - [Story source]()