Sequoia Capital· @sequoia · X·· 1 天前
事实摘要
Ricursive 发布 Ripple-65B Preview，这是其从零预训练的 65B 稠密 LLM。该模型已在 Hugging Face、GMI Cloud 和 OpenRouter 上线，官方称其在智能体编程任务上超过 Muse Spark Open-816B、Beam-501B、Inkling-975B 和 Nemotron 3 Ultra-550B 等更大模型。红杉资本转发了该发布并给出访问链接。
正文 · 原文
Access @RicursiveAI 's model here:
HuggingFace: https://huggingface.co/ricursiveAI/Ripple-65B-PreviewGMI
Cloud: https://console.gmicloud.ai/model/llm/ripple-65b-preview/abfc023e-35b7-4e9e-bb97-cef5db95f171
OpenRouter: https://openrouter.ai/gmicloud/ripple-65b-preview
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://www.ricursive.com/ripple在 X 查看被引用的帖子
来源：Sequoia Capital · x.com