[Feature] Support multiple AI providers (OpenAI, Anthropic, open-source) in addition to Gemini

[iris-i]

Feature Description

ai_call.py currently hardcodes Gemini as the only provider and only has an option to add Gemini models as a .env variable. This is a proposal to expand AI integration so that users can input their API key and select a model from their preferred AI provider.

Detailed Feature Description

Expand ai_call.py beyond its current Gemini-only implementation so it supports multiple AI providers

Suggested providers to support:

• Anthropic
• OpenAI
• Generic options that support self-hosted/open source models like Ollama

Input method:

• People using the web interface, get prompted to put in their API key and model.
• People running the scripts in the terminal can add their API key and models to their env file.

The tool should continue to fall back to Gemini if no API key or model is specified to retain backward compatibility.

Problem and Motivation

ai_call.py is hardcoded to Google Gemini, with no way to select or configure any other model provider. Users who don't have a Gemini API key, prefer a different provider for cost, quality, or compliance reasons, or want to run an open-source/local model instead of a hosted API currently cannot use this tool. Locking the project to a single vendor limits who can adopt it and makes it harder to evaluate or switch providers as needs change.

Potential Benefits

• Easier to select models based on cost and performance
• Easier to compare output across different models
• Enables open-source or local model usage to make the tool easier to adopt for users with budget, offline or privacy constraints.
• Removes the dependency on a single AI vendor.

Target Audience

• End Users
• Developers
• Administrators
• Other

Possible Implementation

No response

Additional Context

No response