A comparison of open large language models, including Llama 4 and Qwen 3.5, is underway. Other models, such as Mistral and possibly Gemma 4, may also be part of the comparison, although some outlets are discussing different models, like Claude Sonnet 5 and TurboQuant. The focus of the discussion appears to be on the capabilities and potential applications of various local LLM models in 2026.