Ricursive Intelligence · Open-weight coding LLM

Ripple-65B Preview

Ricursive Intelligence, an AI lab working on chip design, has released Ripple-65B Preview, a general-purpose variant of the proprietary model behind its AI chip design platform. It is a dense Transformer pretrained on 12 trillion tokens with context up to 1M tokens. The company reports 80.52 on TerminalBench 2.1, which it describes as a U.S. open-weight state of the art. It compares this with Reflection Beam 501B (80.1), Meta Muse Spark Open 816B (76.4), Thinking Machines Inkling 975B (63.8) and NVIDIA Nemotron 3 Ultra 550B (56.4). Ricursive says the full R&D cost was $32M, and that the score came from 1.4M RL rollouts, roughly 70x fewer than Reflection Beam. Architecture choices include 80 layers, d_model 5120, gated attention, global NoPE and value residuals, with 40% MFU during training. The company says its model ladder reaches similar code performance to the Olmo 3 ladder with 16x less compute at 4B scale. The model is available on Hugging Face, Together AI, GMI Cloud, OpenRouter and Fireworks; the text does not state a licence.