Feature Request

We request the integration of the Ollama embedding model, enhanced vision capabilities, and the ability to upload multiple images or files in chat for a seamless user experience. Requested Features: Ollama Embedding Model Support Enable support for Ollama embedding models to enhance AI-powered search, retrieval, and contextual understanding. Improve document and knowledge-based interactions using advanced embedding techniques. Vision Capabilities Introduce support for vision-based models to process and analyze images. Allow AI to extract insights from images, detect objects, and provide relevant responses. Multi-Image and File Upload in Chat Implement functionality to upload multiple images or files simultaneously within chat. Ensure seamless interaction where AI can analyze multiple inputs and provide contextual responses.

Generate videos using text, images or video clips with Replicate API or Runway API.

Allow end users to create their own prompt templates

If you develop and apply MCP Host and Client, you can solve many problems automatically. Please develop it.

I find this API much better for chat than Serper, especially in terms of cost. https://platform.openai.com/docs/guides/tools-web-search?api-mode=chat

Just that... md files support

1.Perplexity (Sonar / Sonar Pro / Sonar Reasoning / Sonar Reasoning Pro) models supported :->Implement features that can be done on the Perplexity site (currently search is done on the Perplexity site) 2.Google Gemini (2.0 Flash / 1.5 Flash / 1.5 Pro) model supported :-> Ability to analyze video (translation, prompted scene finding, see features supported by Google Ai studio) 3.Fine Tuning Models feature :-> Ability to easily select and use your own LLM models that i have trained. 4.Expert Chat bot: The ability to embed a large number of pdf, docx, etc. files and store them in Vector DB and ask questions and get advice as needed.

The feature refers to improving system outputs by enabling Internet access for custom models (on/off). When this feature is enabled, the system can retrieve real-time information from the Web, improving the accuracy of responses. By integrating the Serper API, which leverages Google's search capabilities, the system can provide powerful, personalized information specific to custom templates. This feature enables more relevant and enriched outputs by accessing a larger pool of online data. On the contrary, deactivating this functionality limits the system to its internal resources, possibly limiting the quality and relevance of the answers.

Enable a process driven template where there are multiple steps that can be handed off between the AI models. Example: user inputs a keyword, first step does research, next step writes a blog, next step creates or sources an image, next step provides meta data