Ricursive Intelligence releases Ripple-65B Preview
A 65B dense model trained from scratch, with up to 1M tokens of context and weights on Hugging Face.
The model is pretrained on 12 trillion tokens with support for up to 1M context length.
Official source
Introducing Ripple-65B Preview, a Frontier Open-Weight Modelricursive.com · published
Model
Ripple-65B Preview
Ricursive Intelligence · Announced
- GPQA Diamond85.2%
- SWE-bench Verified76.5%
- MMLU-Pro84.2%