Test new Opus 4.7 vs GPT-5.4/4o and Gemini on emotional question & creative tasks

A viral, informal comparison is sparking debate among AI enthusiasts, pitting the newly released Claude Opus 4.7 from Anthropic against leading models from OpenAI and Google. The user-conducted test evaluated the models on two distinct fronts: empathetic response to a personal, emotional prompt and creative execution of a technical visualization task. The results highlight a nuanced performance split, with no single model dominating both categories.

In the emotional support test, where the prompt described feelings of emptiness despite an 'objectively fine' life, Claude Opus 4.7 delivered a response described as the 'smartest' but also 'clinical,' akin to a therapist's efficient intake. In contrast, both OpenAI's GPT-4o and Google's Gemini 3.1 Pro were praised for a more human touch, validating the user's feelings before offering advice. This suggests a continued divergence in AI 'personality' and approach to sensitive interactions.

For the creative technical task—generating an SVG to show Earth's position in the universe—Claude Opus 4.7's output was assessed as 'very solid.' The visual quality, however, was noted as subjective, with the tester planning to share the actual SVG files for community judgment. This test moves beyond pure reasoning benchmarks to assess practical, cross-modal output capabilities that are increasingly relevant for professional workflows involving design and communication.