Thanks a lot for making this package, I am keen to switching to this for a production app.

Many of the language model providers that AnyLanguageModel already supports have strong multimodal capabilities:

MLX: Supports vision-language models like Qwen2-VL, and other multimodal models
OpenAI: GPT-4o, GPT-4o-mini, and GPT-4-turbo all support vision

Currently, users cannot leverage these vision capabilities through AnyLanguageModel, which limits the library's usefulness for multimodal applications.

Use Cases:

• Image captioning and description
• OCR and document understanding