ResearchSep 17, 2026 · 18:528 sources
Xiaomi Livestreams RL Training for 1 Trillion Parameter Model at $493,000 a Day
From huggingnews.com
Xiaomi is livestreaming the Reinforcement Learning training run of MiMo-V2.6 Pro, a model with 1.02T total parameters and 42B active parameters, via a public dashboard. The stream shows a daily operating cost of $493,000 and a $70,000 cost per step for the Pro version, while a corresponding MiMo V2.6 Flash model with 309B total parameters and 15B active parameters is being trained at a cost of $247,000 per day. Xiaomi's team plans to open source the technical details over the next several weeks, with the final models hosted on Hugging Face.
Read the sourceSign in to suggest
Sources
+ 7 more
- SUPPORT@gauravisnotme“They are at about 5T/18T tokens in a month”x.com
- SUPPORT@madiator“Those behind MOPD are livestreaming a very expensive training run!”x.com
- SUPPORT@eliebakouch“literally livestreaming the RL training run of Mimo V2.6 Pro (1T, 42B active) and Flash (309B, 15B active)”x.com
- SUPPORT@eliebakouch“Mimo V2.6 Pro (1.02T total, 42B active param) 493k$ / day”x.com
- SUPPORT@omarsar0“$1M+ so far”x.com
- SUPPORT@natolambert“One of the coolest at-scale RL resources made public yet!”x.com
- SUPPORT@andrewcurran_“We believe RL is one of the most scalable and efficient paths toward self-improvement.”x.com
#rl#training#open-source#huggingface#compute#scaling#xiaomi#llmMarkdown
Content history (1)
- 2026-10-04 · Title · vi · Wording fixXiaomi Livestreams RL Training for 1 Trillion Parameter Model at $493,000 a DayXiaomi phát trực tiếp huấn luyện RL cho mô hình 1 nghìn tỷ tham số, với chi phí 493.000 U…
One ranked edition, every morning
The stories that matter today. Free, cancel anytime.
Telegram @aidr_today