标签：#AI

OpenAI is sunsetting its AI-powered browser after less than a year. But it's moving some agentic browsing features to its desktop app and a Chrome extension.

Should Anthropic trust Elon Musk to host its models? With about $40 billion in revenue at stake, Musk insists that the company can.

The AI ROI debate has returned and the numbers are even bigger, as are, perhaps, the consequences.

Sandboxing Untrusted Code: Meet CubeSandbox As AI agents become capable of writing, compiling, and running code dynamically, a major security issue has surfaced: how to run this code safely . If a coding agent runs a malicious script or makes an error, it could access files on the host computer or break the entire server. Standard software containers are not always secure enough to prevent escape. CubeSandbox is an open-source, high-performance sandbox service designed specifically to solve this problem. Developed by Tencent Cloud and written in Rust, it provides isolated, secure, and ultra-fast environments for running code generated by AI. What is CubeSandbox? CubeSandbox is a lightweight virtualization system. It spins up a tiny, isolated "virtual machine" for each AI agent task. This ensures that the code runs inside its own virtual bubble, completely separated from the main server. Key Features 1. Hardware-Level Isolation Unlike standard Docker containers that share the same kernel, CubeSandbox uses KVM (Kernel-based Virtual Machine) and RustVMM to give each sandbox its own dedicated Guest OS kernel. This prevents untrusted code from breaking out of the container and accessing your primary server. 2. Under 60ms Cold Starts Traditional virtual machines take seconds to boot. CubeSandbox starts in under 60 milliseconds . This speed is crucial for real-time AI agents that need to execute code instantly. 3. High Density (Low Memory) Each sandbox instance has a memory overhead of less than 5MB . This allows developers to run thousands of concurrent, fully isolated sandboxes on a single physical machine without running out of RAM. 4. Drop-in E2B Replacement For developers currently using E2B (the popular cloud sandboxing SDK), CubeSandbox is fully API-compatible. You can migrate your setup to local hosting by simply changing an environment variable, saving you massive cloud subscription fees. How to Get Started Developers can deploy CubeSandbox locally or in a cluster

From Optimization to Protection: Adding a Security and Governance Agent to Your Snowflake Multi-Agent Team (Part 3) In Part 1 , we built an Admin Agent for usage and cost visibility. In Part 2 , we added a Cost Optimizer Agent and an Orchestrator that routes questions to specialists. Now we close the loop with the third specialist: a Security and Governance Agent . This turns your assistant from "what happened" and "what to optimize" into a full team that also answers "what is risky right now". By the end of this post, you will have: A Security and Governance Agent with focused security tools Security semantic views mapped to natural language Orchestrator routing across Admin, Cost Optimizer, and Security agents A practical triage workflow for failed logins, privilege risk, and unauthorized access Why Add a Security Specialist? The first two agents are strong for operations and spend, but security requires a different lens: Access control and role hygiene Failed login patterns and anomaly detection Unauthorized access attempts Inactive users with active privileges Compliance-friendly audit summaries Could one large agent do everything? Sometimes. But specialized agents are easier to maintain, safer to evolve, and easier to test. Final Team Architecture User Question (natural language) | Orchestrator Agent / | \ Admin Cost Security Agent Optimizer Governance Agent \ | / Unified Response Role of each specialist Admin Agent: usage, credits, storage, operational metrics Cost Optimizer Agent: idle compute, rightsizing, optimization opportunities Security and Governance Agent: roles, privileges, failed logins, unauthorized access, audits The Security Pattern (Same Foundation as Parts 1 and 2) Step 1: Base Views Create security-focused views over SNOWFLAKE.ACCOUNT_USAGE , including: Role hierarchy and privilege grants Failed login attempts and anomaly severity Excessive or unused privileged access Unauthorized access attempts User and role audit summaries Network policy ac

Rebranded Codex promises independent workflows that can run "for hours if needed."

One creative studio for images, video, music, audio, editing, upscaling, and motion control. 50+ models, one credit balance. This is mostly about how we built it and what went wrong along the way. The problem (from a dev perspective) The models are good now. That's not really the issue anymore. The issue is everything around them. Different providers, different UIs, different billing. No shared history across modalities. No easy way to go from "generate image" to "animate it" to "add music" to "upscale" without opening four tabs. We wanted one place where you could actually finish something. What the product is Picturesque has a few main pieces: Studio - tabs for image, video, audio, edit, motion control Projects + Explore - save your work, browse what other people made Workflows - node canvas where you chain models together and run the pipeline in one go Director - an agent that plans multi-step creative work, quotes credits, and runs generations for you The studio covers a lot on its own. 4K images, cinematic video with audio, Suno music, ElevenLabs TTS, Topaz upscaling, motion control, talking avatars. The annoying engineering showed up once we tried to make all of that feel like one product instead of a folder of integrations. Stack (kept boring on purpose) Frontend is React, TypeScript, Vite, React Router. Backend is Node + Express. Socket.IO for real-time updates. Supabase for Postgres and auth. S3-compatible storage for outputs and uploads. For the actual model calls we built a service layer that normalizes inputs, maps our internal model IDs to provider APIs, and handles retries/errors in one place. Media stuff runs through FFmpeg and Sharp. Nothing fancy. When you're wiring up dozens of models with different schemas and pricing rules, you don't want your infra adding more chaos. We also refactored the backend out of a single 7,700-line server.js into routes + services. Painful refactor. Would do it again immediately. The unglamorous part: 50 models, one UI

This is a cross-post — the original (and any updates) live at broke2builtai.com . The first time I watched Claude Code reach for a skill I hadn't told it to use — read a folder, run the script inside it, and hand back the finished thing — the difference from a slash command finally landed. A slash command waits for you to type it. A skill waits for the situation . Claude decides. That one shift is the whole feature, and building one takes about five minutes once you know where the file goes. Here's the entire thing end to end, including the one gotcha that decides whether your skill ever actually fires. What a Skill actually is A Skill is a folder with a SKILL.md file inside it. The Markdown holds instructions; the YAML frontmatter at the top holds a name and a description . That description is doing the most important job in the whole file: Claude reads it to decide, on its own, whether the current task warrants invoking the skill. Nothing else you write matters if the description doesn't get you picked. That's the mental model to hold onto: a custom slash command is a prompt you trigger by typing /name ; a skill is a procedure Claude triggers when the context matches. Same reusable-instructions idea, opposite trigger. Where the file goes Two locations register, exactly like commands and subagents : Project skill — .claude/skills/<skill-name>/SKILL.md inside the repo. Committed, so your whole team gets it. Personal skill — ~/.claude/skills/<skill-name>/SKILL.md in your home directory. Follows you across every project on your machine. Each skill is its own folder, and the folder name should match the name in the frontmatter. A loose SKILL.md sitting somewhere else won't be picked up. The minimum viable skill Create the folder and the file: .claude/skills/pytest-runner/SKILL.md Then write the two-part file — frontmatter, then body: --- name : pytest-runner description : " Run, generate, or debug pytest tests for this project. Use when the user asks to run the test su

Introduction Chrome's Built-In AI APIs allow applications to perform selected AI workloads directly within the browser. Unlike traditional AI integrations, developers do not need to deploy or operate model infrastructure. This guide walks through the major APIs currently available. Getting Started: API Availability and Chrome Flags Chrome's Built-In AI APIs are at different stages of maturity. Some APIs are available in stable Chrome, while others remain experimental. The required setup therefore depends on the API you want to test. Available in Chrome Stable The following APIs are available in stable Chrome on supported desktop devices: Language Detector API Translator API Summarizer API These APIs do not require experimental flags for normal use in supported Chrome versions. The Prompt API has different availability requirements depending on whether it is used from a web page or a Chrome Extension. Check the current Chrome documentation for the environment you are targeting. Experimental APIs The Writer, Rewriter, and Proofreader APIs remain experimental and may require developer trials, origin trials, or Chrome flags for local development. Because these APIs are evolving, refer to the official Chrome documentation for the current setup requirements rather than relying on a static list of flags. Engineering recommendation: Use feature detection and availability() checks at runtime rather than relying on Chrome version numbers or assuming that a particular flag is enabled. Language Detector API Use cases: Dynamic localization Query routing Analytics Content classification Example const detector = await LanguageDetector . create (); const result = await detector . detect ( " Bonjour tout le monde " ); console . log ( result ); Architecture Notes Low latency Task-specific model Suitable for client-side execution Complete runnable example: Language Detector API on GitHub Gist Translator API Use cases: Localization Offline translation International applications Example

The most interesting model comparison of 2026 isn't a benchmark table. It's a product exec quietly changing the question everyone asks about models — and getting a completely different ranking as a result. Claire Vo (founder of ChatPRD, host of the How I AI podcast) ran a head-to-head between OpenAI's new GPT-5.6 lineup (Soul / Terra / Luna) and Anthropic's Claude Fable and Sonnet. The result was an upset: the most theoretically intelligent model, Claude Fable, lost to the one she could actually collaborate with, GPT-5.6 Soul. Here's what that upset actually reveals. She killed "vibes" — then bet 70% back on her own taste Tired of vibe-checking, Vo built a real benchmark across the work she does every day: writing PRDs, prototyping apps, debugging multi-step code, and talking to an agent. Scoring had two layers — an LLM-as-judge (she picked the harshest judge, GPT-5.5) and her own hand-graded "taste test," where she clicked through every artifact and wrote notes. Then the key move: she weighted the final score 70% her taste / 30% the machine. "It's my show. I trust my own taste more." That's the first insight. Benchmarks are getting more rigorous, but the final call is still human taste. The point of blind testing isn't to replace taste — it's to force it to be honest . Cover the labels, react to the work itself, then put your judgment back at the center. Theoretically brilliant vs. practically effective On raw intelligence, Fable is elite. But Vo's verdict is the sharpest line on models I've seen this year: Fable is theoretically hyper-intelligent. Soul is practically effective. She describes Fable as "an engineer who has never met a human." Precise to the point of pedantry — it scores every risk, hardens every edge. In one case it hardened a tool-calling loop so tightly that only one specific model could run it at all. It optimized itself into a corner. Soul's edge was the opposite: it gets out of its own head. Same stuck problem — she moved it to Codex, said "sto

Over the past week, two narratives have been colliding everywhere I look. On one side, there's panic. AI is expected to replace marketers, engineers, and entire categories of knowledge work almost overnight. On the other, there are quieter but far more consequential signals: enterprise teams discovering their AI infrastructure is burning through API budgets far faster than expected. This isn't because the underlying models are weak, but because the systems built around them are fundamentally inefficient by design. These aren't separate stories. They're the same failure showing up in different places. A conversation with another developer made that gap visible in real time. He argued that auditing a 150,000-line codebase requires feeding the entire repository into a model in one single, massive pass. It's still a common assumption in mainstream tech: that an LLM works like a giant biological brain that you must fully load with raw text before it can begin to think. But that assumption is already outdated. Modern AI systems don't scale through brute-force context. They scale through structure. And that shift changes everything. Key takeaways Bigger context windows did not solve AI. Treating a frontier model as a monolithic processor that re-reads an entire system on every query is wasteful, dilutes attention, and hides bugs under raw volume. ARC-AGI-3 makes the gap stark: frontier models scored under 1% on interactive reasoning tasks that untrained humans solve at nearly 100%. The gap is architecture, not memory. The teams pulling ahead treat the model as one narrow component inside a larger system: intelligent routing, task decomposition, retrieval, and only the minimum necessary context. The next advantage is not the biggest model or the longest prompt. It is the system designed around the model. Prompting was the first generation; systems architecture is the next. The Myth of the Infinite Context Window When context windows expanded into the hundreds of thousands o

OpenAI's long-awaited super app is here.

OpenAI is already shutting down ChatGPT Atlas, its browser that could do tasks for you on your behalf, less than a year after launching it. Atlas was announced in October, but as part of its wave of news about ChatGPT Work today, the company confirmed that it will be "sunsetting" Atlas and is targeting an […]

Preclinical trial is testing the feasibility of humanoid robots in surgery.

You can see if ads on Google Search, Google Discover, and YouTube were made or edited using AI from a new section in Google's "My Ad Center," as reported earlier by TechCrunch. The update, announced on Thursday, adds a "created or edited with AI" label under the "how this ad was made" tab. Users can […]

If Google's AI tools are used, it will automatically be noted in My Ad Center.

Meta's pitch to users is Spark's ability to handle large agentic workloads, fix bugs, and help with large code migrations — the kind of automation that enterprises are increasingly turning to AI companies to provide.

News publishers say OpenAI hid tools and datasets that could identify copyrighted journalism in ChatGPT outputs, escalating their lawsuit with a new motion for sanctions.

canonical_url: https://quantumflow-ai-ecosystem.vercel.app/blog/26-ai-models-compared-2026-cost-guide date: 2026-07-09T10:00:00Z If you're building an AI-powered application in 2026, you have a problem: there are too many models to choose from. OpenAI has GPT-4o. Anthropic has Claude 3.5 Sonnet. Google has Gemini 1.5 Pro. Meta has Llama 3.1. And then there's DeepSeek, Mistral, Cohere, and a dozen others. Most developers solve this by defaulting to GPT-4o for everything. It's the safe choice — powerful, well-documented, and reliable. But it's also expensive: $2.50 per million input tokens, $10.00 per million output tokens. If you're processing 10 million tokens a day, that's $75+ per day, $2,250+ per month. But here's the secret: most of your requests don't need GPT-4o. In this guide, we'll compare 26 AI models across three dimensions — cost, quality, and speed — and show you how intelligent routing can cut your AI bill by up to 90% without changing a single line of your application code. The 2026 AI Model Landscape The AI model market has fragmented into three tiers. Understanding these tiers is the foundation of any cost optimization strategy. Tier 1: Sovereign Local Models (Free, Priority 100-110) These models run on your own hardware (or your users' hardware) via runtimes like Ollama. They cost $0 per token. They're sovereign — no data leaves your infrastructure. They're fast (no network round-trip). And they're getting remarkably good. Model Parameters Context Best For Cost Llama 3.1 70B (Local) 70B 128K Complex reasoning, code $0 Llama 3.1 8B (Local) 8B 128K General chat, fast responses $0 Mistral 7B (Local) 7B 32K Efficient European-language tasks $0 DeepSeek Coder (Local) 6.7B 16K Code generation & completion $0 GLM-4 9B Chat (Local) 9B 128K Bilingual (EN/ZH) chat $0 Llama 3.2 3B (Local) 3B 128K Edge devices, mobile $0 Llama 3.2 1B (Local) 1B 128K Ultra-lightweight tasks $0 CodeLlama 7B (Local) 7B 16K Legacy code tasks $0 GLM-4V 9B Vision (Local) 9B 128K Loca

A new $20 billion valuation would be a giant step up from the $10 billion valuation it reached in October.