Hasty Briefsbeta

Show HN: Extending LLM SVG generation beyond pelicans and bicycles

15 days ago
  • #SVG Generation
  • #AI Benchmark
  • #Creative AI
  • Benchmark compares 6 leading AI models on SVG generation from natural language prompts.
  • Models tested: Claude Sonnet 4.5, Grok Code Fast 1, Gemini 2.5 Pro, DeepSeek V3.2-Exp, GLM-4.6, Qwen3-VL-235B-A22B-Thinking.
  • Test involves 30 creative SVG generation prompts.
  • Inspired by Simon Willison's pelican-riding-a-bicycle benchmark.