标签：#GUI

"The reality is, when you're optimizing for production, you start looking at a price/performance," Guillermo Rauch tells TechCrunch.

Whether you’re looking to boost your productivity or your Fortnite stats, these are the top keyboards for the job.

Top up your Qi2 Android phone or MagSafe iPhone with a magnetic wireless charging stand, pad, car charger, or power bank.

Everyone needs a safe way to listen to music on outdoor runs. We’ve found the bone conduction headphones to grab on your way out the door.

Avoid phone FOMO with our favorite smartphones that aren’t officially sold stateside but are available in markets like the UK and Europe.

Lighten your pack this summer—skip the sleeping bag and carry one of our favorite ultralight backpacking quilts instead.

Announced at WWDC 2026, the latest SwiftUI release brings a new Document protocol for efficient disk access and snapshot-based updates, along with improved APIs for reordering items in lists, grids, and sections. In addition, it expands presentation features, such as swipe actions on any view, better AsyncImage caching, and lazy state initialization for Observable types to boost performance. By Sergio De Simone

When going abroad, the right plugs are essential to keep your gadgets charged. These are my favorite travel adapters and chargers.

With these high-tech automatic litter boxes, gone are the days of scooping and smells. Welcome to the future.

The original Ninja Slushi has been replaced! The new best slushie machines chill faster and make soft serve.

WIRED has tested 100-plus bed-in-a-box mattresses for a week each. Our top pick, the Helix Midnight Luxe hybrid, is the best bed you can buy online.

An unadorned ebike is a blank canvas. Here, get tips for maximizing its cargo-hauling and person-carrying capabilities.

After testing each of the MacBooks myself, here's my honest advice on which is right for you.

Discover the best Bluetooth speakers of all shapes and sizes, from waterproof clip-ons to a massive boom box.

I found the best protein powders that won’t make your morning smoothie taste like drywall.

How to Use Chinese LLMs Without a Chinese Phone Number If you've tried signing up for any Chinese AI service, you've seen the same message: Please enter your phone number (+86) to receive a verification code. This single requirement blocks most overseas developers from accessing some of the best-performing and most cost-effective LLMs on the market. This guide covers every workaround I've found — from least to most practical. The Problem China's major AI labs produce world-class models: DeepSeek — DeepSeek V4-Pro matches GPT-4o within 3-5% on coding benchmarks Qwen (Alibaba) — Qwen 3.7-Max beats GPT-4o on long-context tasks (256K tokens) GLM (ZhipuAI) — GLM-4.5 is competitive with Claude for reasoning tasks Baichuan — Strong for Chinese-language generation But every single one requires: A +86 Chinese phone number for registration Alipay or WeChat Pay for billing Chinese-language documentation Method 1: Virtual Chinese Phone Numbers (Fragile) Services like SMS-activate and 5sim offer temporary Chinese phone numbers for ~$1-2. The problem: Chinese providers have gotten aggressive about flagging virtual numbers. Your account gets banned within days. You lose any balance you've added. ❌ Not recommended — too unreliable for production use. Method 2: Third-Party Gateway Services (Recommended) The most practical solution is a gateway that handles the China-side complexity for you. These services: Maintain their own Chinese accounts and infrastructure Register with real Chinese business entities Handle Alipay/WeChat billing on their end Expose everything through a standard OpenAI-compatible API What this means for you: Sign up with email (no phone number needed) Pay via Stripe or PayPal Get a standard API key Use the OpenAI Python/Node.js SDK as-is Migration example (Python): # Before — can't access Chinese models at all # client = OpenAI(api_key="...") # Only works for OpenAI # After — full access to Chinese models client = OpenAI ( base_url = " https://api.tokenmaster.com

Your face called, and it’s low-key offended you might trust TikTok more than WIRED.

I Am Frankelda, A.I. Artificial Intelligence, and From Russia With Love are among the films deserving of your eyeballs this month.

Why AdGuard Home Overtook Pi-hole Last month, while attempting to add ad filtering to the internal network of a production ERP system, the Pi-hole configuration escalated to 85% CPU usage within an hour, causing DNS responses to lag. AdGuard Home resolved the same scenario with 3% CPU and an average latency of 15 ms , which is why it has dethroned Pi-hole. In the following sections, I detail the architecture, performance, security features, and my real-world deployment experience with both products. While they may seem similar at first glance, the fundamental differences directly impact network stability and management overhead. How AdGuard Home Works AdGuard Home is designed as a fully modular DNS forwarder that supports DNS‑over‑HTTPS (DoH) and DNS‑over‑TLS (DoT). Clients first send queries over 53 UDP/TCP or 443 DoH; AdGuard caches the query, checks it against local blacklists, and then forwards it to an upstream DNS service based on preference. # /etc/AdGuardHome.yaml (partial) bind_host : 0.0.0.0 bind_port : 53 upstream_dns : - https://1.1.1.1/dns-query - https://9.9.9.9/dns-query blocking_mode : default blocked_response_ttl : 300 $ dig @127.0.0.1 example.com +short 93.184.216.34 $ curl -s -H "Accept: application/dns-json" "https://adguard.example/dns-query?name=ads.google.com&type=A" | jq . { "Status" : 0, "Answer" : [] , "Question" : [ { "name" : "ads.google.com." , "type" : 1 } ] } Why is it so fast? Cache-first strategy : The initial query goes to an upstream DNS, but subsequent identical domains are returned directly from RAM cache. Parallel upstreams : Since multiple DoH endpoints are tried concurrently, the primary response time drops to an average of 15 ms. Advanced blocklist engine : Thanks to a combination of regex-based filtering and Bloom filters, thousands of ad domains are eliminated in a single query. This architecture, when running as a systemd-based service, shows only 12 ms CPU consumption in systemd-analyze blame output; Pi-hole showed 150 ms

I was convinced these devices were just clutter. I was proven wrong.