AI News, 15 November 2026

Sunday 15 November

Huawei Cloud launches Pangu 2.1 Pro API at 3.2 yuan per million output tokens

AI-generated summary Huawei Cloud added the openPangu-2.1-Pro text generation model to its model-as-a-service platform on November 15. Huawei lists standard input at 0.8 yuan per million tokens and output at 3.2 yuan per million for the API.

Earlier:

Claude Opus 5.5 designs a 200-part analog synth from scratch, and it actually works

AI-generated summary Electronics blogger cpldcpu used Claude Code with Anthropic's Claude Opus 5.5 to build a complete analog synthesizer. The board has around 200 discrete components and the AI designed and simulated the circuit, laid out the PCB, wrote the firmware and generated the order files.

Also covered byexplainx.ai
Earlier:

South Korea to review all 2026 national qualification exams for AI cheating

AI-generated summary South Korea will review all national qualification exams held in 2026 for suspected artificial intelligence-assisted cheating following an emergency presidential meeting. The Ministry of Employment and Labor and the Human Resources Development Service of Korea announced plans to widen the inquiry after the issue emerged during a tax accountant exam.

Earlier:

Wotton Kearney Expands Harvey Legal AI Across Asia Pacific

AI-generated summary Harvey announced that Wotton Kearney is deploying its legal artificial intelligence platform across 13 offices in Australia, New Zealand, Singapore, and Thailand. Wotton Kearney says the firmwide expansion follows a pilot that achieved a 94 percent adoption rate and generated over 30,000 queries.

Also covered bySecurities.io
Earlier:

I asked Claude Code, Google Antigravity, and ChatGPT Codex to build the same game, and only one made something I wanted to keep playing

AI-generated summary The author tested the agentic capabilities of Claude Code, Google Antigravity, and ChatGPT Codex by asking them to build the same 3D time-trial browser racing game. After evaluating the generated games, the author reported that only the output created using Claude Code delivered a truly engaging experience that kept them playing for over forty minutes.

Earlier:

Microsoft to enable Claude by default in British Copilot environments

A computer screen with a sound wave on it

AI-generated summary Microsoft will automatically enable access to Claude models in British Copilot environments on December 9, 2026, unless administrators adjust their settings. In this configuration, documentation indicates that data related to the use of Claude models is processed in the United States.

Earlier:

Cheapest AI APIs Compared: GPT-6, Claude, Gemini, DeepSeek and Qwen Prices (November 2026)

AI-generated summary The Agentic Post published a comparison of AI API prices per million tokens for models including GPT-6 Sol and Luna, Claude 5.5, Gemini, DeepSeek, Grok, Mistral, and Qwen in November 2026. BenchLM listed the cheapest capable models at about 0.10 dollars per million input tokens for OpenAI and Anthropic variants, and about 0.03 dollars for Alibaba's Qwen3.7 Flash.

HP ZGX Fury Packs 748GB RAM and GB300 at $132,000 Price

HP ZGX Fury Packs 748GB RAM and GB300 at $132,000 Price

AI-generated summary HP introduced the HP ZGX Fury G1n AI Station starting at $132,000, featuring Nvidia's GB300 Grace Blackwell Ultra Desktop Superchip and 748GB of memory in a liquid-cooled chassis. HP states that the system pairs 252GB of HBM3e with 496GB of LPDDR5X to let developers fine-tune language models up to 100 billion parameters locally.

Earlier:

OpenAI Business Hub and Business Agents beta

OpenAI Business Hub and Business Agents beta

AI-generated summary OpenAI opened Businesses in ChatGPT in beta on November 10, 2026, introducing Business Hub alongside Business Agents for eligible US ecommerce and online service businesses. OpenAI says Business Hub lets eligible companies claim their organization, update details after verification, and provide product feeds.

Also covered bySI Agent News
Earlier: