Ling 3.0 Flash VL Debuts on OpenRouter with Vision and Visual Agent Capabilities
Key Info
Ling 3.0 Flash VL is now live on OpenRouter, powered by Novita Labs. It is a 124B-parameter Mixture-of-Experts model with 5.5B active parameters, adding native image and video input, visual agent capabilities, hybrid instant/reasoning modes, and tool calling.
Highlights
- Sparse 124B MoE architecture with only 5.5B active parameters, balancing capability and efficiency.
- Native image and video input support plus visual agent features.
- Hybrid instant and reasoning modes, with tool calling for agent workflows.
- Available through OpenRouter's unified API, making it easy to integrate.