I see in the description that only the openAI models (and local models) are available for the analyst feature. Given the release of deepseek-r1 and how much cheaper it is compared to o1 in terms of performance; maybe you would consider implementing support for other models as well.