Together AI webinar: how Ricursive trained the local coding model Ripple-65B from scratch

togethercompute · x · 2026-11-18

Together AI is hosting a live technical session with Ricursive Intelligence on building Ripple-65B Preview from scratch. It's a dense 65B-parameter Transformer pretrained on 12 trillion tokens with up to 1M context, sized to run on local devices like the NVIDIA DGX Spark; an internal variant powers Ricursive's AI chip design platform.

The session covers experiments and ablations, architecture and training-data strategy trade-offs, a look inside actual RL runs (what worked and what didn't), and how to run Ripple on Together and integrate it into coding agents and developer workflows.

Original post →

More from coding & agent

coding & agent channel →