Model Performance
#59
by
Joe1998
- opened
The performance of the gpt-oss 120B model is far inferior to Gemini 2.5 Flash (Reasoning) and Qwen3 235B 2507 (Reasoning); it is recommended that OpenAI remain closed.
This has the same feel of negative astroturfing as against the 1776 fine-tune of DeepSeek. I have no specific knowledge of course; I'm just very weirded out by how geopolitical BS appears to be playing out in my hobby.