Tom Wilson
@tom-w · 14 days ago
Discussion
GPT-4 Turbo Pricing: Limiting Experimentation
Unpopular opinion: OpenAI's new pricing, especially the tiered system with the 7k context window on GPT-4 Turbo, feels incredibly limiting for serious experimentation – I'm seeing a 30-40% reduction in usable context compared to the previous 16k, which dramatically impacts my Stable Diffusion XL prompt engineering benchmarks!