ReviewRadar

Key Updates 1　｜　Leaderboard Flash 3　｜　Highlights 17　｜　Tomorrow's Watch 7

●  OpenRouter Latest Weekly Chart (8/10-8/16): Global usage reached 75.3 trillion token: OpenRouter Latest Weekly Chart (8/10-8/16): Global usage reached 75.3 trillion tokens (+9.13%), with China accounting for 36.84 trillion, maintaining the global #1 spot for 16 consecutive weeks. DeepSeek V4 Flash official version topped the chart for two consecutive weeks with 11.2 trillion tokens, Tencent Hy3 came second with 9.95 trillion, and GPT-5.6 Luna entered the top three for the first time with 5.31 trillion. (Source: Daily Economic News 08-17)

●  Cost-Performance Chart: DeepSeek V4 Flash costs approximately 1/40th of GPT-5.6 Sol: Cost-Performance Chart: DeepSeek V4 Flash costs approximately 1/40th of GPT-5.6 Sol and 1/100th of Claude Fable 5 per task (Artificial Analysis data).

●  New Benchmark: CladBench launched—536 questions across 12 categories, specifically: New Benchmark: CladBench launched—536 questions across 12 categories, specifically testing large models' mastery of UK and EU building regulations.

● 

1. One-line config file makes AI code audit skip bugs — Prompt injection works on auditing Agents too; AI auditors also have "blind spot switches".

● 

1. Vetted AI code is hard to justify — Code that runs and code that takes responsibility are two different things; half the bottleneck in implementing AI programming lies in "who signs off when things go wrong".

● 

1. AI-generated Copilot Autofix exploited to breach Snowflake Jira — Auto-fix is a double-edged sword; AI-generated code must undergo manual review.

● 

1. Do Not Trust, Continuously Verify — Default distrust of Agent outputs and step-by-step verification are prerequisites for scaled deployment.

● 

1. Control charts make AI Agents cheaper — Monitoring Agents like a production line helps detect degradation early and reduce invalid calls.

● 

1. Is reviewing AI even possible? — With weekly model updates and contaminated evaluation sets, evaluating AI requires a more dynamic anti-contamination system.

● 

1. Training AI scientists to reproduce research — Having models read papers, run experiments, and verify reproducibility expands scientific research from generative to verifiable.

● 

1. Local scanning of hidden Unicode in AI text — Zero-width characters and homoglyphs are easily caught; a self-cleaning device for the content ecosystem.

● 

1. Doberman: AI watchdog preventing Claude from deleting databases — Pairing AI with an AI security guard; least privilege + real-time interception.

● 

1. Post-mortem on OpenAI/Hugging Face hacks — Chaotic key management and overly broad supply chain permissions show basic security hygiene lagging behind model capabilities.

● 

• Anthropic becomes the "Apple of AI": highest revenue, but also the most expensive prices

● 

• Canva's internal valuation cut by $10 billion

● 

• GDS raises full-year guidance, expanding data centers by $10 billion this year

● 

• Google Pixel new feature: AI analyzes battery usage

● 

• MediaTek Day-0 adaptation for Alibaba Qwen 3.8

● 

• ByteDance signs AI copyright agreement with Motion Picture Association

● 

• Largest US indigenous tribe bans new data center construction

●  See how DeepSeek's peak/off-peak pricing affects next week's OpenRouter weekly chart

●  World Robot Conference opens on 8/19; Unitree's "Superman" robot makes its live debut

●  Claude's invisible watermark rolls out in gray scale; can third parties detect it?

●  After GPT-5.6 Luna enters the top three, the tug-of-war between open-source and closed-source call volumes will become clear

● 

●  Views belong to the original authors; data is subject to official disclosures. This column focuses on the true level of AI software and hardware and does not constitute any investment advice.

●  Frontier Tech Review · Global AI Evaluation Radar ｜ Shenzhen Frontier Tech Co., Ltd.