←搜索结果

按匹配度 · 共 340 条结果 清除搜索

星期四

  1. The Rundown AI精选

    Z AI 确认神秘模型 Ox Alpha 即 GLM-5.3-Flash

    Z AI 确认匿名免费模型 Ox Alpha 是 GLM-5.3-Flash,并公开模型权重。该模型登顶 OpenRouter,使用量为第二名 DeepSeek 的两倍;Z AI 称免费开放期间的服务完全运行于中国制造的芯片,Artificial Analysis 给出的单任务价格为 $0.045,较同档模型低 10 倍。

    ibe - Google’s speech-to-text AI that edits out filler words 🧠 Qwen3.8-Flash-Next - Alibaba's new open model previewing its Qwen4 design OpenAI shared a post-mortem calling July’s Hugging Face hack a loss-of-control “warning shot,” keeping its biggest frontier training run on hold as it builds automatic shutdown for rogue agents. Bill Gates warned that the world has “no plan” for the AI transition, proposing a tax on AI tokens and robots and “Human Reserved” jobs set aside for people. OpenAI rol

    匹配字段正文

星期二

  1. Hugging Face精选

    IBM 发布 Granite 4.2 语言模型,详解其构建流程

    IBM 发布 Granite 4.2 系列语言模型,包含 3B、8B、30B 三个尺寸,均可输出思维链并支持 thinking 与非 thinking 模式,另设低推理预算的 low-effort 模式。8B 和 30B 额外经过智能体 RL 阶段,学习调用工具、编辑运行代码、操作终端和联网搜索,全部模型支持原生工具调用,可通过 OpenAI 兼容接口接入智能体框架。模型基于 Granite-4.1 base 后训练,SFT 数据约 720 万条、约 100B token,RL 采用异步 GRPO 分阶段课程,并已发布 FP8、NVFP4、MXFP4 和 GGUF 量化版本。

    real-environment trajectories. The Staged Curriculum Each stage is a separate RL run with a single objective and its own reward signal. When it finishes, its policy is exported to Hugging Face format and becomes the base model for the next stage, so the pipeline is a sequence of warm-starts: The 8B and 30B models follow the full ladder. The 3B model takes a shortened path: foundational RL and alignment, without the agentic block. Reward Signals A stage is defined mostly by how it is rewarded . A

    匹配字段正文

星期一

  1. The Rundown AI

    AI 前沿的神秘挑战者:Ox Alpha

    匿名模型 Ox Alpha 免费登陆 OpenRouter,支持 1M-token 上下文和多模态输入,面向编程与持续智能体任务。其 DeepSWE 子集完整测试得分为 63%,接近 Fable 5 且每项任务消耗的 Token 更少;早期线索指向智谱 AI,但开发者身份尚未揭晓。

    Listing Sonar Vortex makes your code conform from the first line, without agents wasting tokens guessing at your architecture. Learn more and drop your token costs by up to 36% . Hugging Face , which hosts over 2M models and 1.5M datasets, is exploring a sale at $13B or more, up from its $4.5B valuation in 2023, Business Insider reported. Nvidia struck a $6B deal to license Poolside’s model-development technology and bring 100+ engineers onto the Nemotron team to build powerful open-weight mode

    匹配字段正文

星期四

  1. 蚂蚁百灵:Developer Blog精选

    蚂蚁百灵开源 Ling-3.0-tiny-base 与 Ling-3.0-flash-base

    蚂蚁百灵正式开源 Ling-3.0-tiny-base 与 Ling-3.0-flash-base 两个 Base Model,并同步开放预训练、中期训练和 WSM 合并三个阶段的权重,共 6 个 checkpoints。Ling-3.0-tiny-base 总参数 7.9B、激活参数 1.3B,Ling-3.0-flash-base 总参数 124B、激活参数 5.1B,可用于持续预训练、监督微调、偏好优化、强化学习后训练与蒸馏。官方说明 Base Model 未完成指令对齐,不建议未经后训练直接部署为面向终端用户的聊天服务。

    h 和 Ling-3.0-tiny,希望为开发者提供覆盖不同规模、不同任务需求的开放模型选择。 开源之后,越来越多开发者开始基于 Ling-3.0 探索自己的应用场景,我们也收到了很多来自社区真实的使用和反馈。大家普遍反馈:希望拥有一个更开放、更具可塑性的 Base Model,以便基于自己的数据、任务和目标开展后训练。 在 Ling-3.0-tiny 的 Hugging Face Community 中,一位开发者直接提出:“Can we get a base model?” 在他看来,Tiny 的模型规模和稀疏架构对预算有限的研究者很有吸引力,适合开展强化学习 rollout 和面向具体场景的后训练;相比已经完成后训练的模型,社区更需要一个能够继续塑造的 Base Model。 今天,我们正式回应这份期待。 01 不只开放最终权重,也开放训练过程中的关键节点 今天,我们将 Ling-3.0-tiny-base 与 Ling-3.0-flash-base 正式开源。除最终 Base checkpoint 外,我们还将同步开放两个模型的预训练和中期训练的权重 checkpoint,共计

    匹配字段正文

星期二

  1. NVIDIA Developer · Generative AI精选

    NVIDIA 发布 Nemotron 3.5 Lightning 开源 MoE 模型,面向长时智能体执行层

    NVIDIA 发布 Nemotron 3.5 Lightning,这是一个 30B 总参数、3B 激活参数的开放 MoE 模型,面向长时运行智能体的高并发低延迟执行任务。模型在预训练阶段内置多 token 预测(MTP)投机解码,并提供 DSpark 与 DFlash 草稿模型,同时发布 NVFP4 量化与 BF16 检查点,可在 Blackwell、Hopper、Ampere GPU 及 DGX Spark 上运行。在 PinchBench 上它以 86% 准确率完成 10000 个任务,速度比 Qwen3.6 35B 快 30%,权重与训练数据按 OpenMDW-1.1 协议开放,并可用 LoRA、SFT 和 NeMo RL 定制。

    30% faster than Qwen3.6 35B at comparable accuracy. Next Steps Try Nemotron 3.5 Lightning on build.nvidia.com to evaluate the model interactively. Download the model weights from Hugging Face for local deployment and customization. Read the Nemotron 3.5 Lightning cookbook for fine-tuning and deployment guides. Powered by NVIDIA Nemotron. AI-generated content may summarize information incompletely. Verify important information. Learn more Long-running AI agents spend most of their time on high-v

    匹配字段正文

  2. Hugging Face精选

    如何使用 NVIDIA Magpie TTS 构建低延迟多语言语音智能体

    NVIDIA Magpie TTS Multilingual 以开放权重支持 12 种语言,并允许开发者在自有基础设施上部署、微调语音和定制发音。NVIDIA NIM 本地测试中,单流首音频延迟为 32–79ms;B200 在 64 路并发时达到 239ms 和 320× 实时吞吐量。

    than it plays back, even under concurrent load. The table above shows Magpie served as the NVIDIA NIM, measured on-prem — the optimized container running on your own GPU. The open Hugging Face checkpoint is the same model and your path for research and fine-tuning; the NIM is the tuned serving stack that produces these production latencies. Both run on hardware you control. Because the model runs on your own infrastructure, you can benchmark performance directly, tune it for your deployment, and

    匹配字段正文

星期一

  1. Hugging Face精选

    Meta 携 Muse Glimmer 回归:本地运行、面向智能体、多模态且开源

    Meta 推出可本地运行的开源多模态模型 Muse Glimmer,由 2B 视觉编码器和 28B 文本解码器组成,总参数量为 30B。它支持图像、无音频视频、多模态工具调用和目标检测,并获得 Transformers、llama.cpp 及基于 Transformers 后端的 vLLM 支持。可选的 DFlash 推测解码器能以增加内存占用为代价加快生成。

    ing , you can test it in the Playground and copy the endpoint URL and model name from the Overview. The deployed model exposes an OpenAI-compatible Chat Completions API. Keep your Hugging Face token in an environment variable, set HF ENDPOINT URL to the URL shown in the Overview without /v1 , and set HF ENDPOINT MODEL to the endpoint's model name. See the Inference Endpoints documentation for configuration, autoscaling, security, logs, and monitoring. Support for Muse Glimmer vLLM with transform

    匹配字段正文

星期三

  1. Hugging Face精选

    Thinking Machines 推出原生多模态开放模型 Inkling

    Thinking Machines 推出 Inkling,一款原生接收文本、图像和音频的开放权重 MoE 模型,拥有 975B 总参数、41B 激活参数和 1M 上下文窗口。其 bf16 与 NVFP4 检查点分别需要 2 TB 和 600 GB VRAM,并已获 Transformers、SGLang、vLLM 和 llama.cpp 支持。

    coding agent harness you can use with different language models. You can use Pi with either an inference engine server endpoint, such as llama.cpp, or with Inference Providers on Hugging Face by adding this to your /.pi/agent/models.json after installation. Then you can start Pi in your project directory by calling pi and you’re good to go! In this demo, we give the model a hard math reasoning problem and it uses tools in pi to solve it. Inkling is focused on broad multimodality reasoning and l

    匹配字段正文

星期三

  1. Google DeepMind精选

    Google DeepMind 发布 DiffusionGemma 实验性开源模型

    Google DeepMind 发布 DiffusionGemma,一个采用 Apache 2.0 许可的 26B MoE 实验性开源模型,探索文本扩散生成,在 GPU 上文本生成速度最高提升 4 倍。该模型不再逐 token 顺序生成,而是同时生成整块文本,面向内联编辑、快速迭代等速度敏感的本地交互工作流。其加速优势主要体现在单加速器、低到中等批大小的本地低并发推理场景,高 QPS 云端服务下并行解码收益递减。

    es your model inference from a single, sequential typewriter to a massive printing press that stamps the entire block of text simultaneously. DiffusionGemma text-to-3D SVG demo by Hugging Face. Step-by-step generation. This means DiffusionGemma's speedup is designed for local and low-concurrency inference. In high-QPS cloud serving, autoregressive models can be deployed to saturate compute efficiently, so DiffusionGemma's parallel decoding offers diminishing returns and can result in higher serv

    匹配字段正文

星期四

  1. LMSYS Blog精选

    SGLang 与 Miles 首日支持 NVIDIA Nemotron 3 Ultra

    LMSYS 宣布 SGLang 与 Miles 在 Day 0 支持 NVIDIA Nemotron 3 Ultra,这是一款面向长时自主智能体的开源前沿推理模型,总参数 550B、激活参数 55B,上下文最长 1M tokens,支持 NVFP4 与 BF16。SGLang 提供高性能推理服务,Miles 支持在 128 张 H200 上以 GRPO 进行强化学习训练,并已验证 dapo-math-17k 上的端到端流程。官方称该模型在智能体生产力、指令遵循和长上下文任务上领先,成本最多节省 30%。

    ing and agentic behavior; Deployment: Open weights, open data, and open recipes for customization and deployment across infrastructure; Supported GPUs: Download model weights from Hugging Face - BF16, NVFP4; Run inference with SGLang using the getting started cookbook; Read the Nemotron 3 Ultra technical report for architecture, training, and benchmark details Installation and Quick Start For an easier setup with SGLang, refer to the Nemotron 3 Ultra getting started cookbook or use the NVIDIA Br

    匹配字段正文

星期五

  1. Mistral AI

    Mistral 发布 Mistral Medium 3.5 与 Vibe 云端远程智能体

    Mistral 发布 Mistral Medium 3.5,这是一个 128B 稠密模型,统一指令跟随、推理与编码,上下文窗口 256k,以修改版 MIT 许可开放权重,可在少至四块 GPU 上自托管。该模型在 SWE-Bench Verified 上得分 77.6%,τ³-Telecom 得分 91.4,并成为 Mistral Vibe 与 Le Chat 的默认模型。同时 Mistral 将编码智能体搬到云端,Vibe 远程智能体可从 CLI 或 Le Chat 启动、并行异步运行,本地会话可迁移上云,完成后可开 PR 并通知;Le Chat 新增 Work 模式(预览),由新智能体驱动多步任务与跨工具流程,敏感操作需人工批准。API 定价为每百万输入 token 1.5 美元、每百万输出 token 7.5 美元,已在 Pro、Team 和 Enterprise 套餐上线。

    and Work mode in Le Chat on the Pro, Team, and Enterprise plans. Through API, it’s priced at $1.5 per million input tokens and $7.5 per million output tokens. Open weights are on Hugging Face under a modified MIT license. It is also available for prototyping, hosted on NVIDIA GPU-accelerated endpoints on build.nvidia.com and as a scalable containerized inference microservice, NVIDIA NIM. Build the future of agentic systems with us. We're hiring across research, engineering, and product to push

    匹配字段正文

星期二

  1. Mistral AI精选

    Mistral AI 发布 Mistral Small 4,统一推理、多模态与编码能力

    Mistral AI 发布 Mistral Small 4,这是 Mistral Small 系列的下一个主要版本,也是 Mistral 首个把 Magistral 的推理、Pixtral 的多模态和 Devstral 的智能体编码能力统一到单一模型中的版本,采用 Apache 2.0 许可。模型为 MoE 架构,128 个专家、每 token 激活 4 个,总参数 119B、每 token 激活 6B(含嵌入和输出层为 8B),上下文窗口 256k,支持文本与图像输入,并可通过 reasoning_effort 参数在快速响应与深度推理之间切换。官方称相比 Mistral Small 3,端到端完成时间减少 40%,每秒请求数提升 3 倍;推理模式下在三项基准上匹配或超过 GPT-OSS 120B,且输出更短。

    Performance Comparison Across Internal Models

    search, and complex reasoning tasks Its open-source license and customizable architecture make it ideal for fine-tuning and specialization. Availability Mistral API and AI Studio; Hugging Face Repository; Developers can prototype Mistral Small 4 for free on NVIDIA accelerated computing at build.nvidia.com, and for production deployment, Mistral Small 4 is available day-0 as an NVIDIA NIM, delivering optimized, containerized inference out of the box. It can also be customized with NVIDIA NeMo for

    匹配字段正文

星期一

  1. 蚂蚁百灵:Developer Blog

    蚂蚁百灵发布并开源 Ling-2.5-1T 即时模型

    蚂蚁百灵发布并开源旗舰级即时模型 Ling-2.5-1T,总参数 1T、激活 63B,预训练语料从前代 20T 扩展至 29T,支持最高 1M token 上下文。模型采用 1:7 的 MLA + Lightning Linear 混合线性注意力架构,引入正确性与过程冗余复合奖励机制,在 AIME 2026 上平均输出约 5890 token 即接近消耗 15k–23K token 的前沿思考模型水平。官方称其在复杂推理与指令遵循上优于 DeepSeek V3.2、Kimi K2.5、GPT 5.2,可适配 Claude Code、OpenCode 等智能体产品,但在复杂智能体交互和长程任务上仍落后于前沿模型。

    rompt:请你写一个 minialipay,前后端分离架构 + 内存数据库存储,支持基础的用户登录、转账等功能,并且有新春特色的集五福功能,欢庆春节主题。 Opencode 接入(skills) 能够调用正确的 Skills,完整执行内部系统开发指令,Agent 自动化完成从基础架构搭建到多品牌主题逻辑的代码实现,将需求描述直接转化为具备导航、搜索及动态换肤功能的可运行原型。 Minecraft 机器人任务执行 精准执行空间构建指令,Agent 自动化完成从底层铺设到顶层收缩的数百次方块放置,将自然语言直接转化为游戏内的实体建筑。(使用的 Agent 框架 6. 局限性与未来计划 Ling-2.5-1T 已实现高吞吐解码与领先的超长上下文处理能力,并初步具备智能体交互能力,为通用智能体时代奠定基础。但在复杂智能体交互和长程任务上,它仍落后于前沿模型。下一版本将聚焦真实世界复杂任务的长程执行与交付效率优化,同时持续提升 Token 效率,以在效率与效果间取得更优平衡。 欢迎通过以下方式下载、访问和使用 Ling-2.5-1T。 🤗 Hugging Face: 🤖 ModelScope:

    匹配字段正文

  2. Qwen Blog精选

    Qwen3.5 发布:首个原生多模态智能体模型 Qwen3.5-397B-A17B

    Qwen 团队正式发布 Qwen3.5 系列首个开源权重模型 Qwen3.5-397B-A17B,这是一个原生视觉语言模型,在推理、编码、智能体和多模态理解等基准上给出完整评测结果。模型采用线性注意力(Gated Delta Networks)与稀疏 MoE 混合架构,总参数 397B、每次前向仅激活 17B,语言与方言支持从 119 种扩展到 201 种。托管版本 Qwen3.5-Plus 通过阿里云百炼提供,默认 1M 上下文窗口,并内置官方工具与自适应工具调用。

    s | Qwen Blog Publication About Try Qwen Chat Qwen3.5: Towards Native Multimodal Agents February 14, 2026 · 196 min · 41706 words · Qwen Team | Translations: 简体中文 QWEN CHAT GitHub Hugging Face ModelScope DISCORD We are delighted to announce the official release of Qwen3.5 , introducing the open-weight of the first model in the Qwen3.5 series, namely Qwen3.5-397B-A17B . As a native vision-language model, Qwen3.5-397B-A17B demonstrates outstanding results across a full range of benchmark evaluatio

    匹配字段正文

星期四

  1. Mistral AI精选

    Mistral AI 发布 Voxtral Transcribe 2,实时转写延迟低于 200ms

    Mistral AI 发布 Voxtral Transcribe 2,包括批量转写模型 Voxtral Mini Transcribe V2 和实时模型 Voxtral Realtime。两者支持 13 种语言;前者提供说话人分离、上下文偏置和词级时间戳,单次可处理最长 3 小时音频,API 价格为 $0.003/min。4B 参数的 Realtime 延迟最低可配置至 200ms 以下,API 价格为 $0.006/min,并以 Apache 2.0 许可开放权重。

    Fleur Voxtral 2

    h a 4B parameter footprint, it runs efficiently on edge devices, ensuring privacy and security for sensitive deployments. We’re releasing the model weights under Apache 2.0 on the Hugging Face Hub. Voxtral Mini Transcribe V2. Average diarization error rate (lower is better) across five English benchmarks (Switchboard, CallHome, AMI-IHM, AMI-SDM, SBCSAE) and the TalkBank multilingual benchmark (German, Spanish, English, Chinese, Japanese). Average word error rate (lower is better) across the top-

    匹配字段正文

星期一

  1. Qwen Blog精选

    Qwen 发布旗舰推理模型 Qwen3-Max-Thinking

    Qwen 团队发布旗舰推理模型 Qwen3-Max-Thinking,通过扩大参数规模并投入大量算力做强化学习,在事实知识、复杂推理、指令遵循、人类偏好对齐和智能体能力上均有提升,在 19 项基准上表现与 GPT-5.2-Thinking、Claude-Opus-4.5、Gemini 3 Pro 相当。模型新增自适应工具调用能力,可在对话中自主调用内置的 Search、Memory 和 Code Interpreter,已在 chat.qwen.ai 上线;同时采用经验累积的多轮测试时扩展策略,在 GPQA、HLE、LiveCodeBench v6 等基准上优于标准并行采样。API 模型名为 qwen3-max-2026-01-23,兼容 OpenAI 与 Anthropic 接口协议,可直接配合 Claude Code 使用。

    Bench 38.2 56.3 51.6 44.1 40.9 Deep Planning 6 44.6 33.9 23.3 21.6 28.7 Long Context AA-LCR 72.7 74.0 70.7 65.0 68.7 We evaluated only on the text subset. ↩︎ We blocked access to Hugging Face and other HLE-related websites to prevent data leakage. ↩︎ For reproducibility of Arena-Hard v2, we report the win rates evaluated by GPT-4.1. ↩︎ We followed the official setting of Tau² Bench with no custom scaffolding. ↩︎ BFCL-V4 is configured with a maximum of 100 interaction turns. ↩︎ Deep Planning is

    匹配字段正文

星期五

  1. vLLM 官方博客

    NVIDIA Nemotron Nano 2 VL 发布并获 vLLM 支持

    NVIDIA Nemotron Nano 2 VL 开源视觉语言模型发布,面向视频理解与文档智能,并获 vLLM 推理支持。模型采用混合 Transformer–Mamba 架构,通过 EVS 剪除冗余视频帧以减少 token。vLLM 夜间版本支持以 BF16、FP8 和 FP4 精度部署 OpenAI 兼容的推理服务器。

    r; Efficient video sampling as token compression module; Hybrid Transformer-Mamba Architecture - Nemotron Nano 2 LLM backbone with reasoning; Accuracy: Download model weights from Hugging Face - BF16, FP8, FP4-QAD; Run with vLLM for inference; Technical report to build custom, optimized models with Nemotron techniques Run optimized inference with vLLM This guide demonstrates how to run Nemotron Nano 2 VL on vLLM, achieving accelerated inference and serving concurrent requests efficiently with BF

    匹配字段正文

星期日

  1. Prime Intellect精选

    Prime Intellect 发布 INTELLECT-2:首个通过全球分布式强化学习训练的 32B 模型

    Prime Intellect 发布 INTELLECT-2,称其为首个通过全球分布式强化学习训练的 32B 参数模型,在动态异构的算力贡献者集群上以完全异步 RL 训练推理语言模型。团队为此从零构建了异步 RL 框架 PRIME-RL、验证推理 worker rollout 的 TOPLOC 和广播策略权重的 SHARDCAST,并修改标准 GRPO 训练配方与数据过滤方法以稳定训练,在数学和编码基准上提升了 QwQ-32B 的表现。模型、代码和数据均已开源,技术报告见 primeintellect.ai/intellect-2。

    INTELLECT-2 Release: The First 32B Parameter Model Trained Through Globally Distributed Reinforcement Learning

    with our code and data, hoping to enable more open research in the field of globally-distributed training Detailed Technical Report: primeintellect.ai/intellect-2; INTELLECT-2 on Hugging Face Chat Interface to try it out: chat.primeintellect.ai; prime-rl: – our async RL framework Paradigm Shift for Distributed Training Test-time compute scaling with reinforcement learning has emerged as a new scaling axis for large language models (LLMs), enabling improvements by allowing models to spend more t

    匹配字段正文

星期四

  1. Mistral AI精选

    Mistral AI 发布 24B 参数模型 Mistral Small 3

    Mistral AI 发布延迟优化的 24B 参数模型 Mistral Small 3,并以 Apache 2.0 许可开放预训练和指令微调检查点。官方称其 MMLU 准确率超过 81%、推理速度达 150 tokens/s,与 Llama 3.3 70B instruct 表现相当且在同一硬件上快 3x 以上。量化后可在单张 RTX 4090 或配备 32GB RAM 的 Macbook 上本地运行,并已通过 la Plateforme 及合作平台提供。

    Up and to the Left

    ilable on la Plateforme as mistral-small-latest or mistral-small-2501. Explore our docs to learn how to use our models for text generation. We are also excited to collaborate with Hugging Face, Ollama, Kaggle, Together AI, and Fireworks AI to make the model available on their platforms starting today: The road ahead It’s been exciting days for the open-source community! Mistral Small 3 complements large open-source reasoning models like the recent releases of DeepSeek, and can serve as a strong

    匹配字段正文

星期一

  1. Prime Intellect

    Prime Intellect 等发布 7B 宏基因组基础模型 METAGENE-1

    USC、Prime Intellect 与 Nucleic Acid Observatory 合作发布 7B 参数自回归 Transformer 宏基因组基础模型 METAGENE-1,在超过 1.5 万亿碱基对的污水宏基因组 DNA 与 RNA 序列上预训练,用于疫情监测、病原体检测和新兴健康威胁的早期发现。模型采用针对宏基因组序列的 BPE 分词,在病原体检测和宏基因组嵌入基准上取得 SOTA,并演示了病原体与异常检测场景。团队称其 512 token 上下文长度限制了复杂序列设计任务,当前版本风险较低,但未来更大模型开源前需严格安全评估。

    METAGENE-1: Metagenomic Foundation Model

    undation model designed for pandemic monitoring, trained on over 1.5T base pairs of DNA and RNA sequenced from wastewater. Paper: metagene-1-paper; Github: github.com/metagene-ai; Hugging Face: huggingface.co/metagene-ai A collaboration between researchers at USC, Prime Intellect, and the Nucleic Acid Observatory. Overview We pretrain METAGENE-1, a 7-billion-parameter autoregressive transformer language model, which we refer to as a metagenomic foundation model, on a novel corpus of diverse meta

    匹配字段正文