搜索“Agent”

一个可继续训练、拓展并适配真实场景的开放底座。 在下表的评测中，模型在代码、复杂推理和长上下文方向表现突出。相比总参数量约为其 2–3 倍的 base model，模型的整体表现仍具优势。 这使 flash-base 适合用于以下方向的继续训练与探索： 面向软件研发和大型代码库的代码模型； 需要多步分析与复杂推理的专业模型； 面向长文档、长流程和长程任务的 Agent 基础模型； 金融、医疗、科研、工业等专业领域的持续预训练与后训练。 Benchmark 描述的是特定设置下的 Base Model 能力，并不等同于聊天体验、Agent 执行效果或生产系统性能。Tiny 和 Flash 在知识、数学等部分评测上也各有能力边界，我们希望社区以此作为选择训练底座的参考。 03 为什么要开放 Base Model？ 真实世界不只有一种任务，也不只有一种正确答案。 真实世界任务往往具有长期目标、专业知识、复杂环境、工具调用和责任边界。医疗、金融、科研、工业和个人助手需要不同的数据、反馈机制和安全约束。一个官方后训练模型可以提供通用能力，却无法替代每个领域基于真实任务进行的训练、对齐和验证。 Ba

匹配字段正文

as surpassed Moonshot AI’s Kimi K3 and on some it’s surpassed Claude Fable 5 or GPT-5.6-Sol. Here’s a more complete comparison: This puts the model more or less at the frontier of agentic coding benchmarks, with only 750B parameters – a third of Kimi K3! The Z.ai blog post is rather straightforward, and starts with a bold sentence: Scaling post-training is all we did for GLM-5.3. GLM-5.3 is the same base model as GLM-5.2 with substantially extended post-training. To risk a broad oversimplificati

匹配字段正文

LI; X; FB Today we’re announcing MAI-Cyber-1-Flash inside of MDASH, our multi-agent vulnerability identification and remediation harness. Together they deliver world-class performance at 50% of the cost of leading models. Progress in AI has been startling and so has the new generation of cyber threats it’s unleashing. Attackers now wield increasingly powerful capabilities, probing an ever-growing mountain of code for just a single weakness that lets them in. As the cost of finding a flaw collaps

匹配字段正文

As our most intelligent workhorse model yet for coding and agents, Gemini 3.7 Flash brings higher instruction adherence, first-pass code accuracy, and high-quality execution for agentic tasks. See how the model handles building a complex 3D web game with 3.7 Flash in @antigravity: 1️⃣ 3D Engine Logic: Generates Three.js boilerplate, camera movement, and collision detection. 2️⃣ Asset Orchestration: Creates PBR-textured materials, custom sprite sheets with Nano Banana, and particle effects. 3️⃣ P

匹配字段正文

Aug 13, 2026 Our most intelligent workhorse model yet for coding and agents. Tulsee Doshi Senior Director, Product Management, on behalf of the Gemini team Your browser does not support the audio element. Today, we’re building on the progress of our widely used Flash series by introducing Gemini 3.7 Flash, our most intelligent workhorse model yet for coding and agents. This release comes just three weeks after Gemini 3.6 Flash , and is a direct result of developer feedback and algorithmic innova

匹配字段正文

ce can be deployed, how often it can be used, and whether it can move from exceptional tasks into daily workflows. We have invested heavily in the training environments needed for agentic coding. Each verified environment is deterministic, executable, and graded by real test suites. This gives the model practice on the kind of multi-step work developers actually do: reading code, editing files, running tests, observing failures, and recovering from intermediate mistakes. Advanced mathematical re

匹配字段正文

Table of Contents Today we are releasing Grok 4.6 together with SpaceXAI. Grok 4.6 builds on Grok 4.5 with a particular focus on long-running agents and more ambitious interactive and visual work. It stays with complex tasks across many steps, whether researching a topic, analyzing information, working across a codebase, or turning an idea into a polished application or work artifact. Grok 4.6 achieves frontier intelligence across several agentic coding and knowledge work benchmarks. It matches

匹配字段正文

Today we are releasing Grok 4.6 . Grok 4.6 builds on Grok 4.5 with a particular focus on long-running agents and more ambitious interactive and visual work. It stays with complex tasks across many steps, whether researching a topic, analyzing information, working across a codebase, or turning an idea into a polished application or work artifact. 0:00 / 0:00 Grok 4.6 achieves frontier intelligence across several agentic coding and knowledge work benchmarks. It matches GPT-5.6 Sol on the Artificia

匹配字段正文

m('facets.search.text-copy'); sessionStorage.removeItem('facets.search.filters'); NVIDIA Nemotron 3.5 Lightning Delivers Fast, Accurate Specialized Task Execution for Long-Running Agents Aug 11, 2026 NVIDIA Nemotron 3.5 Lightning is an open 30B mixture-of-experts model with 3B active parameters optimized for high-volume, low-latency execution in long-running AI agents. The model includes speculative decoding with multi-token prediction baked in during pretraining, plus DSpark and DFlash draft mo

匹配字段正文

2026 NVIDIA Nemotron 3.5 Lightning is now available on Ollama, and it runs completely on your own device. It’s a 30 billion parameter (3B active) open model from NVIDIA built for agents that stay running: gathering context, calling tools, and working through multi-step tasks. Nemotron 3.5 Lightning is made for agentic tasks such as reading a file, calling a tool, sorting a result, and retrying something that failed. Most of these steps don’t need a large model. At 3B active parameters per token

匹配字段正文

s.search.text-copy'); sessionStorage.removeItem('facets.search.filters'); Aug 10, 2026 Muse Glimmer is a 30B open-weight dense model with a 120K+ context window built for local AI agentic work. The dense architecture activates every parameter for each token, delivering reliable instruction following, long-context coherence, predictable latency, and fewer failure modes for multi-step agentic workloads. NVIDIA Tensor Core architecture accelerates Muse Glimmer to over 20K tokens per second on a sin

匹配字段正文

st 10, 2026 Muse Glimmer, Meta’s newest open model and the first released by Meta Superintelligence Labs, is now available on Ollama. It’s a 30B multimodal model purpose-built for agent workloads that run locally with a 128K+ context length, released under the Apache 2.0 license. With Ollama, you can now use Muse Glimmer to power coding agent applications such as Claude Code, Codex, Pi and more, as well as long-running personal assistants such as OpenClaw and Hermes. On Apple Silicon, it is avai

匹配字段正文

Join us for our inaugural conference, Forge 2026 PUBLISHED 8/10/2026 This model was built with agents in mind. It is designed for "always-on" agents rather than just simple chat interactions. It’s capable of managing many sequential tool calls over multiple turns. It has the ability to recover from failures and retries rather than halting when a tool call fails or returns something unexpected. Muse Glimmer is a 30B dense model made up of 52 transformer layers, grouped-query attention with 32 que

匹配字段正文

Ling-3.0-flash 采用 124B 总参数、5.1B 激活参数的 MoE 架构。它保留了千亿参数模型的知识与能力容量，但在生成每个 Token 时，只激活其中约 5.1B 参数，从而降低单次推理的计算开销。 这让 Ling-3.0-flash 不只是一款“跑得快”的模型，也能够真正进入不同规模的业务系统，提供三种不同的落地选择： API 调用，让 Agent 应用快速上线： 面向希望快速接入、无需自建推理服务的开发者，Ling-3.0-flash 可通过云端 API 直接调用； 单机私有化，让数据留在本地： 面向数据不能出域的企业和团队，MXFP4 与 INT4 版本可在单台 NVIDIA DGX Spark 上完成端到端推理； 高性能部署，释放极致单请求速度： 面向单请求时延敏感的高性能服务，在指定 GPU 测试配置下，平均输出速率突破 1100 tokens/s。 开发者既可以直接调用模型，也可以自由验证、部署和定制模型能力，将 Ling-3.0-flash 接入自己的业务系统、开发工具链和 Agent 工作流。 从云端 API 到单台 DGX Spark、AMD 395

匹配字段正文

ent , end = "" , flush = True ) answer content += delta . content For more information, please visit the API doc . Coding Assistants Qwen3.8-Max integrates seamlessly with popular agent frameworks and coding assistants: Claude Code Qwen APIs support the Anthropic API protocol, enabling direct use with Claude Code : npm install -g @anthropic-ai/claude-code export ANTHROPIC MODEL = "qwen3.8-max" export ANTHROPIC SMALL FAST MODEL = "qwen3.8-max" export ANTHROPIC BASE URL = https://dashscope-intl.al

匹配字段正文

ed to be very efficient with reasoning tokens relative to its predecessor. In production settings, this means that tool calls are snappier, usually executing before the end of the agent's first sentence. Reasoning efficiency Relative reasoning tokens per response (P50). Lower is better. Conversational Capability We've trained Grok Voice Think Fast 2.0 to be a better conversationalist. By using extensive reinforcement learning to push the model towards patterns we see in real human conversation,

匹配字段正文

Kimi K3 is now live on @togethercompute! Happy to have Together AI as our day0 launch partner, giving developers immediate access to K3 through high-throughput inference optimized for coding agents and production workloads. A huge thank-you to the Together AI team for moving fast and making this launch possible!

匹配字段正文

Releasing the model weights and technical report of Kimi K3. Kimi K3 is our most capable model: a 2.8T MoE model with native visual understanding and a 1M-token context window. New model architecture: 2.5x the intelligence per unit of compute, not just more params. Alongside Kimi K3, we're opening up more of the stack behind it — high-performance attention kernels, MoE communication library, and infrastructure for running agent environments at scale.

匹配字段正文

odels. K3 is a mixture-of-experts transformer: 2.8T total parameters, 16 of 896 experts active per token, a 1M token context window, and native vision. It's built for long-horizon agentic work, and Moonshot put that to the test internally, handing an early version most of the team's kernel optimization work during development. Kimi Delta Attention holds down the cost of attention as sequences grow, and Attention Residuals let deeper layers reach back to earlier attention outputs rather than only

匹配字段正文

s on track through long multi-step analyses.Alfredo Andere CEO Claude Opus 5 came out ahead of every model in its family on our internal evals. It isn’t just better on our hardest agentic coding tasks, up 22% over Opus 4.7, it’s steadier, with far less variance run to run. For the millions of builders on Lovable, that consistency is the whole game. Reliable results, build after build.Fabian Hedin Co-Founder Claude Opus 5 is the biggest leap in the Opus family since 4.5. On the same full-stack ap

匹配字段正文