Together AI 推出了 Inkling-Small，一个由 ThinkyMachines 开发的开放权重多模态模型。该模型拥有 2760 亿参数，其中 120 亿活跃，提供跨文本、图像和音频的原生推理能力，并支持高达 100 万个 token 的上下文窗口。Inkling-Small 在性能上可与更大的 Inkling 模型相媲美，但成本和尺寸显著降低，使其适用于编码、代理任务和通用多模态应用。 AI
影响 为开发者提供了一个面向生产的多模态模型路径，有可能加速其在编码和代理应用中的采用。
排序理由 该集群宣布了一个特定 AI 模型在平台上的可用性，属于工具或产品集成类别。
在 Mastodon — fosstodon.org 阅读 →
AI 生成摘要 · Google Gemini · 来自 4 个来源。 我们如何撰写摘要 →
Together AI gives developers a high-throughput production path for running Inkling-Small on @NVIDIAAI Accelerated Infrastructure across multimodal, coding, and agentic workloads. Start building: https://t.co/cDoeF5OLHM
Together AI gives developers a high-throughput production path for running Inkling-Small on NVIDIA Accelerated Infrastructure across multimodal, coding, and agentic workloads. Start building: https://t.co/BOGutQYlCo
NVIDIA has released Molt, a PyTorch-native agentic reinforcement learning framework designed for building and deploying autonomous AI agents at scale. The framework aims to simplify the development of multi-step automation systems. https://www. marktechpost.com/2026/08/01/nv idia…
Physical AI infrastructure is shifting from compute-centric to layered systems addressing perception, motion, and failure recovery. NVIDIA leads with accelerated computing and simulation, but the stack now requires interdependent platforms for data, validation, and continuous lea…