此语言仍有 24 条内容或界面任务待翻译;未完成部分显示中文原文。

Insights & News

AI、算力、具身智能机器人与智能硬件的产业动态,以及选型与采购指南。

今日更新0 篇Updates This Week23 篇Published147 篇订阅周报

2026-10-23 · 本站更新

大模型与软件Industry Updates

Moore Threads Completes Day-0 Adaptation of the MiniMax M3.1 Flash Large Model

On 10 23, MiniMax open-sourced its multimodal large model MiniMax M3.1 Flash. On the same day, Moore Threads' flagship AI training-and-inference compute card MTT S5000 and its proprietary MUSA software stack completed Day-0 adaptation of the model. With the adaptation timing synchronized with the model's open-source release, the model can run on Moore Threads' domestic GPU platform as soon as it is open-sourced, though the source provides no details on performance, deployment scale, or ecosystem.

观点网 · 约 4 分钟

2026-10-22 · 本站更新

大模型与软件Industry Updates

Ant Group's Open-Source Ling-3.1-flash Deployment Compilation: 560B Hybrid MoE, H200/GB300 Verified Recipes, and SGLang Launch Essentials

Ling-3.1-flash, open-sourced by Ant Group's inclusionAI, is a hybrid-architecture MoE language model with 560B total parameters and 25B activated per token, with FP8 quantized weights. According to a CSDN blog compilation, as of 2026 10, 15, the SGLang official deployment page verified two recipes: 8×H200 and 4×GB300; the official 54 safetensors shards total approximately 573.8 GB. Deployment requires multi-GPU resources and a dedicated image; the context length has three definitions — 1M nominal, 256K on the official service, and 262,144 in the weight config — and first deployments should start with a short context.

CSDN 博客(作者:aidoudoulong,七牛云行业应用) · 约 7 分钟

大模型与软件Industry Updates

DeepSeek Open-Sources DeepInfer Inference Framework and Releases deepseek-recipe Component

On 20 10, DeepSeek announced it would fully open-source its internal inference framework DeepInfer under the Apache-2.0 license, along with the MIT-licensed token encode/decode component deepseek-recipe. DeepInfer has already supported DeepSeek V4.1-Flash large-scale production inference and currently supports NVIDIA SM100 and Huawei Ascend 950 platforms, but it is an offline inference engine without online serving or a standalone KV cache system, and supports only the DeepSeek series of models.

鲲鹏昇腾开发者社区(CSDN) · 约 6 分钟

2026-10-21 · 本站更新

大模型与软件Industry Updates

StepFun Open-Sources Step 5 Preview: 600B-Parameter MoE with 1M Context, Built for Long-Horizon Agents

On 15 10, 2026, StepFun open-sourced the Step 5 Preview model weights, available in both FP8 and BF16 formats. The model uses a sparse MoE architecture with 600B total parameters and 27B activated parameters, contains 92 Transformer blocks, and supports up to 1M tokens of context; the company says it targets long-horizon Agent tasks in software engineering and professional knowledge work. Commercial use requires reviewing the additional conditions of the StepFun Model License 1.0.

七牛云 · 约 10 分钟

大模型与软件Industry Updates

MiniMax Releases M3.1 Flash: Focused on Agentic Task Performance and Cost

MiniMax officially released MiniMax M3.1 Flash. Officially, it previously ran under the anonymous model name Space Bunny, cumulatively processing over 200 trillion tokens. The new model focuses on agentic tasks, coding, Computer Use, and cybersecurity, with officially published benchmark results including AutomationBench-AA, BankerToolBench 76.2, DeepSWE v1.1 73.7%, OSWorld V2 61.7, ALE 29.6, and CyberGym 82%. Officials claim nearly 3x faster online inference speed and a 70% cost reduction versus the previous generation, priced at input $0.15, output $0.60, cached input $0.015 (per 1M tokens). The model is now available via MiniMax API, M Plan, and MiniMax Code, with weights to follow.

同一事件另有 1 篇报道

MiniMax Research · 约 5 分钟

大模型与软件Industry Updates

阶跃星辰开放 Step 5 Preview 权重,600B 参数稀疏 MoE 模型上线 Hugging Face 与 ModelScope

阶跃星辰于 2026 年 10 月 15 日正式开放 Step 5 Preview 模型权重,含 BF16 与 FP8 两个版本,并附技术博客、开源开发工具及部署指南,权重已上线 Hugging Face 和 ModelScope。该模型面向长程 Agent 任务设计,采用稀疏 MoE 架构,总参数 600B,每 token 激活 27B,支持 100 万 token 上下文。据 OpenRouter 截至 10 月 15 日数据,其 Token 用量周榜及日榜均列全球第二。

科创板日报 · 约 2 分钟

2026-10-20 · 本站更新

大模型与软件Industry Updates

Meta 开放 Muse Spark Open:816B 参数 MoE 模型的 Hugging Face 申请与 SGLang/vLLM 部署路径

Meta 于 2026 年 10 月 15 日发布开放权重模型 Muse Spark Open,总参数 816B、每 token 激活 42B,上下文窗口 1,048,576 token,原生支持文本、图片、音频、视频输入。本文依据官方部署文档整理权重体积(BF16 约 1.65TB、NVFP4 约 537GB)、B300 GPU 与 Apple MLX 参考硬件、SGLang/vLLM 启动命令、reasoning_effort 参数及 muse-model-license 的 MaaS 营收条款。

CSDN博客(七牛云行业应用) · 约 4 分钟

大模型与软件Industry Updates

Qianwen Office E-commerce Edition Launches, AI Agent Boosts Merchant Operations Efficiency

On 10 14, 2026, Qianwen Office CEO Chen Yusen disclosed at the e-commerce edition launch event that more than two months after launch, the product's paid user share in the e-commerce sector exceeded 20%. A merchant case presented at the event showed that after adopting the AI agent, new store preparation time was compressed from one month to 7 days, initial investment dropped from over 1 thousand yuan to a few hundred yuan in Token costs, and daily operations work takes only 2 to 3 hours.

软盟资讯 · 约 5 分钟

2026-10-16 · 本站更新

Industry InsightsIndustry Updates

Meta releases Muse Wonder model series, focusing on agent tasks and professional-domain capabilities

According to Meta's official press release, Meta launched two models, Muse Wonder and Muse Wonder Full, on 10/13, describing them as the company's most capable models to date, achieving industry-leading performance on personal and professional agent tasks. The new models strengthen professional-domain knowledge in law, finance, and healthcare as well as cutting-edge coding abilities, and introduce two new features, Active Memory and Active Focus. They are available starting today via Meta Model API and Muse Code, with subscribers getting first access.

腾讯新闻 / PANews · 约 3 分钟

2026-10-04 · 本站更新

硅基之家示例内容 · 约 1 分钟

Product Selection Assistant

让需求变成可修改的筛选条件

依据已录入参数生成条件,结果可查看和修改。帧率与性能需核对实测,不能仅凭算力数值保证。