In a dynamic week for artificial intelligence, Google launched Gemini 3 on November 18, emphasizing multimodal reasoning and visual comprehension, which reflects a strategic shift in its AI approach. Following closely, Anthropic introduced Claude Opus 4.5, claiming it surpassed human performance in internal assessments. Both companies reduced operational costs significantly, with Anthropic cutting prices by 67%, making Claude Opus 4.5’s processing cost $5 per million tokens, while Google priced Gemini 3 Pro at $2 for reading and $12 for generation, leveraging new TPU infrastructure.
Gemini 3 excels in handling diverse formats such as text, images, and video, posting over 90% on advanced scientific reasoning benchmarks. In contrast, Claude Opus 4.5 enhances coding capabilities and long-term reasoning, outperforming Gemini in engineering tasks. Each model addresses industry needs: Gemini for visual tasks and Opus for structured logic. Both signify a move towards advanced AI capable of managing multistep workflows, highlighting evolving economic efficiencies in AI deployment.
Source link
