搜索“Agent”

Listen to this blog post Try it out Listen to this blog post Today, we're excited to announce a step change in xAI's Voice Agent capabilities: Introducing grok-voice-think-fast-1.0 — our new flagship voice model. This new model excels at complex, ambiguous, multi-step workflows across customer support, sales, and enterprise applications. It is especially well-suited for high-stakes scenarios that demand precise data entry and high-volume tool calling to address the user's request. See it in acti

匹配字段正文

rameter multimodal model at the scale the community has been asking for most. Still supporting both multimodal thinking and non-thinking modes, Qwen3.6-27B delivers flagship-level agentic coding performance, surpassing the previous-generation open-source flagship Qwen3.5-397B-A17B (397B total / 17B active MoE) across all major coding benchmarks. As a dense architecture, it is straightforward to deploy without MoE routing complexity, making it an ideal choice for developers who need top-tier codi

匹配字段正文

our next proprietary model: Qwen3.6-Max-Preview . Compared to Qwen3.6-Plus, this preview release brings stronger world knowledge and instruction following , along with significant agentic coding improvements across a wide range of benchmarks. As a preview, the model is still under active development — we are continuing to iterate and expect further gains in subsequent versions. Qwen3.6-Max-Preview is the hosted proprietary model available via Alibaba Cloud Model Studio , featuring: improved agen

匹配字段正文

ows. It cuts the friction from those multi-step tasks so developers can stay in the flow and focus on building.Mario Rodriguez Chief Product Officer Based on our internal research-agent benchmark, Claude Opus 4.7 has the strongest efficiency baseline we’ve seen for multi-step work. It tied for the top overall score across our six modules at 0.715 and delivered the most consistent long-context performance of any model we tested. On General Finance—our largest module—it improved meaningfully on Op

匹配字段正文

Qwen3.6-35B-A3B: Agentic Coding Power, Now Open to All | Qwen Blog Publication About Try Qwen Chat Qwen3.6-35B-A3B: Agentic Coding Power, Now Open to All April 15, 2026 · 8 min · 1639 words · Qwen Team | Translations: 简体中文 QWEN STUDIO HUGGING FACE MODELSCOPE DISCORD Following the launch of Qwen3.6-Plus , we are excited to open-source Qwen3.6-35B-A3B — a sparse yet remarkably capable mixture-of-experts (MoE) model with 35 billion total parameters and only 3 billion active parameters. Despite its

匹配字段正文

he global market. In the Vision domain, we selected Qwen 3.5 27B, Qwen 3 VL 235B-A22B, Qwen 3 VL 32B, Claude Sonnet 4.5, and GPT-5-mini as comparison models, while in the Text and Agent domains, we selected the previous version, EXAONE 4.0 32B, and the K-EXAONE 236B model as comparison models. 3.1 Performance evaluation of the Vision field in the industrial field For vision-language models (VLMs) to be adopted in real-world industrial settings, they must go beyond simple image recognition and po

匹配字段正文

We’re releasing SWE-1.6, our latest model built for software engineering agents, and we’re making it generally available in Windsurf. SWE-1.6 is optimized for both intelligence and model UX. Moreover, it is industry-leading in both speed (up to 950 tok/s) and cost (free tier for the next 3 months). Last month, we released SWE-1.6 Preview, which improved on SWE-Bench Pro by more than 10% compared to our previous model SWE-1.5 while being post-trained on the same pre-trained model. SWE-1.6 was pos

匹配字段正文

securely create your own custom voice in Microsoft Foundry with just a few seconds of audio. MAI-Voice-1 can transform how easily developers can build voice experiences and voice agents – at high quality and high speed. The model can generate 60 seconds of audio in just a single second, and highly efficient GPU usage delivers that quality and speed affordably. Hearing is believing, so experience it for yourself with Copilot Audio Expressions or Copilot Podcasts. MAI-Image-2 has turbocharged ima

匹配字段正文

ning SOTA performance across 25 languages. Outstanding performance in noisy environments Benchmarks are only part of the story. When it comes to production use cases such as voice agents, meeting transcription, and call center analytics, audio is rarely clean. MAI-Transcribe-1 was built with challenging recording conditions in mind, reliably handling background noise, low-quality audio recordings, and overlapping speech. Cafe scenario TRANSCRIPTION: Hey, so I was hoping to change my flight, if t

匹配字段正文

Qwen3.6-Plus: Towards Real World Agents | Qwen Blog Publication About Try Qwen Chat Qwen3.6-Plus: Towards Real World Agents April 2, 2026 · 127 min · 26994 words · Qwen Team | Translations: 简体中文 QWEN CHAT DISCORD Following the release of the Qwen3.5 series in February, we are thrilled to announce the official launch of Qwen3.6-Plus. Available immediately via our API, this release represents a massive capability upgrade over its predecessor. Most notably, we have drastically enhanced the model’s

匹配字段正文

Short Video Analysis Storyline 00:00.000 – 00:04.300 The clip opens with a hand-held, wide-angle selfie shot. A Black man in his twenties, sporting short curly hair and a bright magenta-yellow-red striped long-sleeve top, fringed magenta skirt, matching leg warmers, and large fluffy magenta ankle attachments, fills most of the frame. Behind him stretches a sun-bleached, sandy clearing under a blue sky mottled with white clouds. Dozens of onlookers—men, women, and children in casual clothes—form

匹配字段正文

y for download. Speech is rapidly becoming a core modality for AI-enabled workloads and automations — from meeting transcription and speech analytics to real-time customer support agents. Our objective was straightforward: push the frontier of dedicated ASR model accuracy under practical conditions. The model was trained from scratch with a deliberate focus on minimizing word error rate (WER), while keeping production readiness top-of-mind. In other words, not just a research artifact, but a sys

匹配字段正文

ation. The model excels in contextual understanding, speaker modeling, and zero-shot cross-lingual voice adaptation, making it ideal for enterprise voice workflows and scalable AI agents. Available via API and Mistral Studio, it offers cost-effective, high-quality voice generation starting at $0.016 per 1k characters. Today we’re releasing Voxtral TTS, our first text-to-speech model with state-of-the-art performance in multilingual voice generation. The model is lightweight at 4B parameters, mak

匹配字段正文

Composer 2 is out! Cursor is an example of a new type of company, not a pure app maker and not a model provider. Our aim is to build the most useful coding agents by combining the best API models and our domain-specific models.

匹配字段正文

of post-training methods, especially with reinforcement learning with verifiable rewards (RLVR) for coding or math, requires huge amounts of generated rollouts, and most recently, agentic workflows, such as Codex, Claude Code, or even OpenClaw, have pushed inference demand through the roof . Despite the clear, growing importance of inference, many linear architectures (including Mamba-2) were developed from a training-first perspective. To accelerate pretraining, the underlying SSM was progressi

匹配字段正文

istral Small family. Mistral Small 4 is the first Mistral model to unify the capabilities of our flagship models, Magistral for reasoning, Pixtral for multimodal, and Devstral for agentic coding, into a single, versatile model. With Small 4, users no longer need to choose between a fast instruct model, a powerful reasoning engine, or a multimodal assistant: one model now delivers all three, with configurable reasoning effort and best-in-class efficiency. Mistral Small 4 is released under the Apa

匹配字段正文

AI agents have proven to be highly capable tools at code generation. Yet, as we push these models to high-stakes domains, ranging from frontier research mathematics to mission-critical software, we encounter a scaling bottleneck: the human review. The time and specialized expertise required to manually verify become the primary impedance of engineering velocity. We envision a more helpful generation of coding agents to both carry out their tasks and formally prove their implementations against s

匹配字段正文

of post-training methods, especially with reinforcement learning with verifiable rewards (RLVR) for coding or math, requires huge amounts of generated rollouts, and most recently, agentic workflows, such as Codex, Claude Code, or even OpenClaw, have pushed inference demand through the roof. Despite the clear, growing importance of inference, many linear architectures (including Mamba-2) were developed from a training-first perspective. To accelerate pretraining, the underlying SSM was progressiv

匹配字段正文

fficient across a wide range of hardware, from flagship GPUs to personal devices like laptops. Both models are already in production. Sarvam 30B powers Samvaad, our conversational agent platform. Sarvam 105B powers Indus, our AI assistant built for complex reasoning and agentic workflows. The Sarvam models are globally competitive for their class. Sarvam 105B performs well on reasoning, programming, and agentic tasks across a wide range of benchmarks. Sarvam 30B is optimized for real-time deploy

匹配字段正文

We chose to evaluate SWE-1.6 on SWE-Bench Pro following OpenAI’s recommendation as the spiritual successor of SWE-Bench Verified. Running bug-free and reproducible evaluations for agents requires care. We manually read hundreds of trajectories and cross-checked against Scale-reported SWE-Bench Pro trajectories when applicable. Some examples of subtle issues we resolved include: dependency issues in agent and grading environment setup, inconsistent handling of timeouts across harnesses, edge case

匹配字段正文