Ricursive Intelligence Releases Ripple-65B Preview, Training Cost $32 Million
Ricursive Intelligence has released Ripple-65B Preview, a 65B dense large model pretrained from scratch, claiming it achieves SOTA on agentic coding tasks among U.S. open-weight models, surpassing larger models such as Muse Spark Open-816B, Beam-501B, Inkling-975B, and Nemotron 3 Ultra-550B. swyx cited its explanation stating that the $32 million is the complete R&D cost from scratch, covering pretraining ladder development, data mixing, evaluation, debugging, synthetic data generation, post-training scheme development, and the final pretraining and post-training runs.
Why read this
AI score 73
Ripple-65B Preview, with a full-process R&D cost of $32 million, surpasses multiple larger open-source models on agentic coding tasks.
Topics
View original postToday we release Ripple-65B Preview, our 65B dense LLM we pretrained from scratch. Ripple sets a new state-of-the-art for U.S. open-weight models in agentic coding tasks, surpassing significantly larger models including Muse Spark Open-816B, Beam-501B, Inkling-975B, and Nemotron 3 Ultra-550B. Learn more: https://t.co/Pe04ZKhkE9
View original
Today we release Ripple-65B Preview, our 65B dense LLM we pretrained from scratch. Ripple sets a new state-of-the-art for U.S. open-weight models in agentic coding tasks, surpassing significantly larger models including Muse Spark Open-816B, Beam-501B, Inkling-975B, and Nemotron 3 Ultra-550B. Learn more: https://t.co/Pe04ZKhkE9
Body
> we developed and trained Ripple-65B Preview for only $32M, which is one-tenth the cost of the previous open-weight SOTA on agentic coding¹
this is like the avengers of open model SOTA
¹ "Note that $32M is our full R&D cost starting from scratch, including pretraining ladder development, data mixing, evals, debugging, synthetic data generation, posttraining recipe development, and the final hero pretraining and posttraining run."