[Enhancement]: OpenRouter Image Generation Tool

[JumpLink]

What features would you like to see added?

Add a dedicated Image Generation tool for OpenRouter that supports all OpenRouter-compatible image generation models, including FLUX.2-Pro, FLUX.2-Flex, Gemini Image Generation, and other models available through OpenRouter.

This tool would provide native support for OpenRouter's image generation API, which uses /v1/chat/completions with modalities: ["image", "text"] instead of the standard OpenAI /v1/images/generations endpoint.

More details

Current Situation:

• OpenRouter supports image generation through models like black-forest-labs/flux.2-pro, black-forest-labs/flux.2-flex, and google/gemini-2.5-flash-image-preview
• These models use OpenRouter's chat completions API with a modalities parameter, not the standard images endpoint
• Existing image tools (OpenAI Image Tools, Flux, DALL-E) are designed for their respective APIs and don't work well with OpenRouter's API structure

Use Case:
Users who want to use OpenRouter's image generation models (FLUX.2-Pro, FLUX.2-Flex, Gemini Image Generation, etc.) through LibreChat's Agents feature currently cannot do so easily, as the existing tools are not compatible with OpenRouter's API structure.

Requirements:

• Support for OpenRouter's /v1/chat/completions endpoint with modalities: ["image", "text"]
• Support for multiple OpenRouter image generation models (FLUX.2-Pro, FLUX.2-Flex, Gemini Image Generation, Riverflow v2, etc.)
• Proper handling of OpenRouter's response format (images are returned in message.images array)
• Support for aspect ratio configuration for models that support it (e.g., Gemini models)
• Integration with LibreChat's existing tool system and authentication

Documentation:

• OpenRouter Image Generation: https://openrouter.ai/docs/guides/overview/multimodal/image-generation
• Supported models: https://openrouter.ai/models?fmt=cards&input_modalities=image

Which components are impacted by your request?

Plugins

Pictures

No response

Code of Conduct

• I agree to follow this project's Code of Conduct

[JumpLink]

I've started implementing this feature! The PRs are currently in draft mode:

• Implementation: feat: Add OpenRouter Image Generation tool #11024
• Documentation: docs: Add OpenRouter Image Generation tool documentation LibreChat-AI/librechat.ai#470

The tool supports multiple OpenRouter image generation models with aspect ratio control. Would love to get your feedback 🚀