We need a 80-160B model urgently. The unified memory device market needs more Models.

The author emphasizes the urgent need for large language models in the 80-160 billion parameter range to support users with unified memory devices and high RAM configurations. Current offerings, such as the 27B Qwen and 31B Gemma, cater primarily to low-capacity machines, leaving those with 80-128GB setups reliant on older or smaller models that may not perform adequately. The proposed models, like GPT OSS 120B V2 and Qwen 3.6, would better leverage the capabilities of high-memory systems, addressing a gap in the market for AI practitioners.