AnalysisAI ModelsJuly 1, 2026
User asks for model to fill 144GB VRAM + 120GB RAM
The user currently uses Qwen3.6 27b and Gemma4 31b, and wants the largest model that fits 144GB VRAM and 120GB RAM. Minimax M2.7 at Q6 (207GB base) is considered tight.
The user currently uses Qwen3.6 27b and Gemma4 31b, and wants the largest model that fits 144GB VRAM and 120GB RAM. Minimax M2.7 at Q6 (207GB base) is considered tight.