feat: Add ModelsLab text-to-image and text-to-video nodes

[adhikjoshi]

Summary

Adds two new ComfyUI nodes for ModelsLab — a developer-first AI API platform providing image/video generation using Flux, Stable Diffusion, CogVideoX and more.

New Nodes

ModelsLab Text-to-Image (category: ComfyCloudAPIs/ModelsLab)

• Supports Flux, SDXL, Realistic Vision v6, DreamShaper 8, Anything v5
• Full controls: width/height, steps, guidance scale, seed, negative prompt
• API key stored in keys/ folder (same pattern as existing nodes)

ModelsLab Text-to-Video (category: ComfyCloudAPIs/ModelsLab)

• Supports CogVideoX-5B and Wan-T2V models
• Handles async processing with automatic polling (up to 5 min)
• Returns VideoFromFile compatible with existing video nodes

Usage

1. Get a free API key at https://modelslab.com/dashboard/api-keys
2. Save as a .txt file in the keys/ folder
3. Add the node to your workflow (find under ComfyCloudAPIs/ModelsLab)

API Reference

• Image Generation API
• Video Generation API

Code Notes

• Follows existing node patterns exactly (key loading, image tensor conversion, etc.)
• No new dependencies — uses requests already in requirements
• Video node handles both sync and async (processing) responses