Ling 2.6 Flash

Ling-2.6-flash is an instant (instruct) model from inclusionAI with 104B total parameters and 7.4B active parameters, designed for real-world agents that require fast responses, strong execution, and high token efficiency. It delivers performance comparable to state-of-the-art models at a similar scale while significantly reducing token usage across coding, document processing, and lightweight agent workflows.