Ricursive releases Ripple-65B Preview, claiming US open-weight SOTA on agentic coding
DZhang50 · x · 2026-11-05
Ricursive Intelligence released Ripple-65B Preview, a dense LLM pretrained from scratch on 12T tokens with 1M context. It scores 80.52 on TerminalBench 2.1, claiming US open-weight SOTA on agentic coding and beating much larger models: Reflection Beam-501B (80.1), Meta Muse Spark Open-816B (76.4), Thinking Machines Inkling-975B (63.8), NVIDIA Nemotron 3 Ultra-550B (56.4). Originally built to power its AI chip-design platform, the small dense model runs locally on DGX Spark; available on Hugging Face, Together AI, GMI Cloud, and OpenRouter, with Fireworks coming. The lab claims frontier open-weight performance at one-tenth the usual cost via hardware-software-model co-optimization.
More from Models
- Morph launches Reflex: first decision model with cached pricing, 10-20x cheaper than Jev — GabGarrett · 2026-11-05
- Hamel Husain: Benchmarks are used like credit scores, so criticism is warranted — HamelHusain · 2026-11-05
- Report: 80-90% of OpenAI research is focused on GPT-7 and beyond — VraserX · 2026-11-05
- Latent Space: weight-space programming replicated across labs, Tencent Hy4 tops agents, Anthropic open-sources FlashProve — Latent Space · 2026-11-05
- Early demos show Kimi K3.1 Pro generating 3D worlds directly — _akhaliq · 2026-11-05
- Ripple-65B team adds hardware verification lead Caroline Trippel amid launch — annadgoldie · 2026-11-05