搜索“Agent”

keeps compute and memory flat as sequences grow. For endpoints on Modal, we took this even further with Day 0 support for a custom DFlash speculator tuned to this model shape. On agentic workloads on 8x B200s, we saw 250 tokens per second per user at 2.5M TPM of per-GPU throughput, 67% faster interactivity than the model's built-in speculative path at matched throughput. Towards local attention: Evolving DFlash speculation for frontier attention models Inkling’s unique architecture moves in a d

匹配字段正文

ugh Observability, lineage and telemetry trace a production output back to the version of the asset behind it, and back to the usage that prompted the last change. The skills your agents run are reachable as MCP servers straight from Studio, so what executes in production is the same governed asset you versioned, not a copy that drifted. You define behaviour, watch it run, and improve it, all against one source of truth. That closed loop is the difference between cataloging your AI and governing

匹配字段正文

, EMEA, and beyond, and a one-month minimum term. Inference spend has become a line item the board asks about. In a not too distant future every knowledge worker will have several agents working on their behalf, but companies are struggling to achieve this without astronomic inference budgets. To succeed companies need to consume a range of models that span the Pareto frontier, matching task difficulty with model quality (and cost). Open weight models occupy an ever-greater fraction of that fron

匹配字段正文

You can now start new agents remotely from ampcode.com anywhere you can run amp: That means, in addition to running agents in orbs, you can now run agents on any machine you want: your laptop, your server, your cloud dev box, your Raspberry Pi. Your lawn mower even, if it has a shell. Enable it by using the command amp: enable remote creation of threads or with the setting: Once enabled, every Amp client you start will accept and run new threads in its working directory. Runner Mode You can also

匹配字段正文

eir flow. There’s no context switching, no manual environment setup, and no permission troubleshooting. More from this author Record, train, and deploy from one place with Strands Agents, LeRobot, and Hugging Face Storage Buckets 34 August 13, 2026 From the Hugging Face Hub to robot hardware with Strands Agents and LeRobot 17 June 17, 2026 Community Edit Preview Upload images, audio, and videos by dragging in the text input, pasting, or clicking here . Tap or paste here to upload images Comment

匹配字段正文

Microsoft Foundry is a platform for building and operating agentic AI applications. Foundry starts with the widest model selection on any cloud — models from Microsoft, OpenAI, Anthropic, Meta, Mistral, DeepSeek, Hugging Face, and others, spanning frontier, open-source, and custom weights — all accessible through a single endpoint and a single set of SDKs in Python, C , JavaScript, and Java. On top of those models sits the Foundry Agent Service : multi-agent orchestration with built-in memory, k

匹配字段正文

runtime, and ships remediation PRs. Security Swarm finds more verified vulnerabilities at 30% lower cost than the nearest comparable alternative. How it works A swarm of parallel agents investigate segments of the codebase. Each agent reasons across files—catching business logic flaws, chained auth bypasses, and cross-service exploit paths. Devin then composes individual findings into full attack paths and reproduces each one in an isolated sandbox to confirm exploitability at runtime. What rea

匹配字段正文

Internally, a significant amount of engineering time is consumed by manually verifying that the code an agent writes works E2E in dev. We sought to change that. We built an internal agent, the Verifier, to take over that step. On every run it deploys the change to a live environment, exercises the affected behavior, and posts what it observed (logs, API responses, and screenshots) back to the PR. It runs on Cosmos, our cloud-agent platform. Why not write more tests? We weighed a maintained suite

匹配字段正文

You can now launch Amp agents remotely in orbs: Orbs are machines where agents can run without supervision. When you create a new thread you get a fresh orb that contains your code, plugins, and tools the agent might need. Orbs are billed by the minute. See the current orb sizes and pricing. They start up quick and go to sleep when you and the agent no longer need them. Remote and Yet So Near Agents in orbs aren't running on your machine, but you can control them just like they were. You can rev

匹配字段正文

dynamic mid-session routing. We welcome you to try Devin Fusion in preview at app.devin.ai/signup. The Trick: Sidekick The key idea behind our architecture is to run two parallel agents: one with a frontier model, the other with a more cost-effective "sidekick" model. Both are fully capable agents with their own toolsets and ability to gather & act on their own context. As the task progresses, the main agent decides which tasks to give the sidekick and which tasks to do itself. Making sidekick

匹配字段正文

June 29, 2026 Gemma 4 is now significantly faster in Ollama 0.31. On Apple Silicon, it generates tokens nearly 90% faster on average across a coding-agent benchmark. The speedup is on by default, and it does not change the model’s output: The speedup comes from multi-token prediction (MTP). Gemma 4 ships with a small, fast draft model that runs alongside the main model and proposes the next several tokens. The main model then verifies that proposal in a single pass and keeps the tokens it agrees

匹配字段正文

omponents, your variables, and your team—so your designs can come to life from day one.Inside Figma; Product updates; Figma Design; Motion; 3D design; Config; News; Figma’s design agent, now with custom tools and greater contextJune 24, 2026By Georgia Rust, Rodrigo DaviesPrompting gets you started. Custom tools, real context, and skills take the design agent further—so what ends up on the canvas actually looks like you.Inside Figma; Product updates; Figma agent; Figma Design; Config; AI; News; C

匹配字段正文

Prompting gets you started. Custom tools, real context, and skills take the design agent further—so what ends up on the canvas actually looks like you. Share Figma’s design agent, now with custom tools and greater context Explore our playground files for the agent and custom tools. The difference between an agent that helps you work more efficiently and one that actually understands how you work is context. When an agent knows your team's way of working it can do more than produce. It can collab

匹配字段正文

(GA) to reuse your connectors in developer interfaces; Connectors in Workflows (Public Preview) allow for uninterrupted long-running tasks powered by all the tools you need Async agents are moving into everyday work. For an agent to be trustworthy and useful inside an organization, it needs real enterprise data: CRM records, repositories, inboxes, knowledge bases. Connecting an agent to that data in a demo is easy. Running it in production is where most setups stall. Production connectivity has

匹配字段正文

AI and ML The Claude in Slack app is dead, long live Claude in Slack Anthropic is killing off its existing Claude in Slack app in favor of an always-on agentic alternative dubbed Claude Tag that will listen and learn from everything it’s given access to. Claude Tag can now join organizational Slack instances “as a team member,” according to Anthropic’s announcement, where it will have access to whatever channels domain administrators decide upon, as well as the tools, data, and codebases they co

匹配字段正文

entire autoscaling, multi-replica service behaves under load. Of course, there is no universal configuration for inference. A low-latency classification endpoint and a multi-turn agentic loop do not want the same serving setup. Modal Auto Endpoints start you at the configuration we’d start from before pulling traces: clean, inspectable, benchmarked, and ready to tune against a workload. Engine-level observability Modal provides metrics (in-Dashboard and via OTEL export) to understand endpoint p

匹配字段正文

Today we are releasing prime-rl version 0.6.0. This version enables us (and you) to train models of trillion-parameter scale on heavy agentic workloads at the highest efficiency. We have been relentlessly optimizing our RL infrastructure to maximize performance on large MoE models, reducing the cost, time and suffering required to post-train OSS models on agentic workflows. We are able to train GLM-5 on SWE tasks at up to 131k sequence length, with sub-5-minute step times and a batch size of 256

匹配字段正文

Today we're introducing /goal in Grok Build. This new mode yields long-running, autonomous execution, helping you hand off larger implementation tasks to the agent. Available now in Grok Build. Install with a single command and sign in with your account: Install Grok Build Upgrade Introducing /goal Most coding sessions require back-and-forth execution and verification. With /goal , the agent continues until a task is completed and verified, whether that means reviewing code, inspecting webpages,

匹配字段正文

ulation at 250 tok/s, using the token timing simulator from our LLM Engineer’s Almanac: Furthermore, they better preserve their acceptance lengths on very long context tasks, like agentic software engineering. Below, we explain why we’re bullish on speculation for LLM inference acceleration — and as part of the whole continuous improvement cycle for AI applications. But first, a tl;dr with the high-level takeaways. To first order, speculative decoding is the only engine optimization that matters

匹配字段正文

s republished it twice as the investigation progressed, incorporating the latest information. With artifacts, team members and stakeholders don’t have to "walk us through what the agent found" because they're all looking at the same view, with the same context. Private to your organization Every artifact is private to its author by default. When you're ready, share it with your teammates and your organization directly from the page. Artifacts are viewable only by authenticated members of your or

匹配字段正文