I built a tool to create LLM Tier Lists based on real tasks

Show HN: I built a tool to create LLM Tier Lists based on real tasks

I built a tool to create LLM Tier Lists based on real tasks operates in show hn launch.

Profile

Hey HN,Following up on my prompt tool. I realized that "prompt engineering" is useless if we don't know which model actually handles the instruction best. We often argue if GPT-5 is better than Gemini 3 Pro, but it usually depends on the specific use case.So I built a module to create Head-to-Head Model Benchmarks (Tier Lists).How it works:Define a specific task (e.g., "Writing a LinkedIn launch post").Run it simultaneously on 14+ models (including the latest GPT-5, Gemini 3, Claude 4.5 families).Compare metrics: Speed, Token usage, and qualitative output.Rank them: Drag and drop models into S, A, B, C tiers based on the results.Why I built this: I found that for coding, Model X might be S-T

Within the show hn launch space, the company focuses on its core offering rather than a broad product line.

The company is listed at the Show HN launch stage of development.

Website: promt.oshn-ai.com/community/004471c6-b508-4ae8-a7cd-20ce6ab4ad65