Full Summary
This Monday morning, multiple sources, including OfficeChai, The Decoder, and Crypto Briefing, confirm Sakana AI has launched its new Fugu AI system, which remarkably matches the performance of top-tier models like Anthropic's Fable and Mythos. This Tokyo-based company, valued at over $2.5 billion, achieves this by orchestrating multiple underlying models through a single API, rather than building one massive model. Blockchain.news and OfficeChai both highlight that this multi-agent approach improves accuracy and robustness in reasoning, coding, and multimodal understanding, offering frontier-level results while mitigating regulatory risks. The Decoder adds that Fugu comes in a base version and a more powerful Fugu Ultra, which performs on par with leading models in coding, reasoning, and science benchmarks. Meanwhile, Google has launched Gemini 3, its new flagship AI model, replacing Gemini 2.5 Flash, according to iNews Zoombangla. It offers significant improvements in speed and reasoning, focusing on "agentic" tasks and excelling in coding. News-Medical reports that Google's AMIE AI system matched primary care doctors in disease management, even outperforming them in treatment recommendations. In other major AI news, Zhipu AI’s market capitalization has topped one trillion Hong Kong dollars after its open-source GLM-5.2 model was released last week. The South China Morning Post and Invezz both note GLM-5.2 now ranks second globally on the Code Arena list for front-end web development, trailing only Anthropic's Fable 5. Interconnects AI and Lapaas Voice emphasize GLM-5.2's open-weights design and its ability to act as an agent, often outperforming leading paid US models on certain tests at about one-sixth the cost of GPT-5.5. Geeky Gadgets and NewsBricks report that OpenAI is set to release ChatGPT 5.6 Pro as early as next week, with a significant enhancement to AI reasoning power and an increased reasoning effort budget. This new version will integrate Playwright for web automation and has a knowledge cutoff of December 2025. This rapid pace of AI development means that advanced, cost-effective AI solutions are becoming more accessible, potentially impacting everything from medical diagnoses to your next software update.