Model Library
Browse and deploy state-of-the-art AI models through the DEVUP Gateway.
Browse and deploy state-of-the-art AI models through the DEVUP Gateway.
Ling-3.0-flash-Fin is the first finance-enhanced model in Ant Group's Ling family, built by continuing the training of Ling-3.0-flash on high-quality financial data with input from financial institutions and domain experts. It carries 124 billion total parameters but activates only 5.1 billion per token — roughly four percent — which keeps inference light enough for the long agent runs financial research actually requires. Across a 256K context it connects retrieval, evidence review, calculation, modelling, and report preparation as one workflow rather than separate tasks, reconciling conflicting figures across annual reports, earnings releases, and filings. Released under the MIT licence, it retains the general reasoning, coding, and mathematics of the model it extends.
