Ling 3.1 Flash scores 41 on the Intelligence Index with just 25B active parameters
As seen on the 24/7 Wall St. homepage on October 6, 2026.
Ant Group's open-weight reasoning model lands at 41 on the Intelligence Index with only 25B of its 560B parameters active, another sign Chinese labs are closing the agentic gap at a fraction of the compute cost.
Ling 3.1 Flash makes large gains in intelligence over its predecessor, scoring 41 on the Artificial Analysis Intelligence Index with particular improvement in agentic capabilities @AntLingAGI has released Ling 3.1 Flash, a reasoning model with 560B total parameters, 25B active https://t.co/p2RgbV8XIV
Continue ReadingShow less
Ant Group's AntLingAGI released Ling 3.1 Flash on October 6, 2026, a reasoning model built on a mixture-of-experts architecture with 560 billion total parameters but only 25 billion active at any given time. The model delivers meaningful intelligence gains while keeping actual compute demand a fraction of what a dense model of comparable scale would require.
Artificial Analysis benchmarked the release at 41 on its Intelligence Index, a notable step up from the predecessor model. The evaluation specifically called out agentic capabilities as the area of sharpest improvement, meaning the model is better at planning, tool use, and multi-step reasoning tasks that matter most for real-world AI deployment.
Agentic benchmarks are increasingly the measure that enterprise buyers and developers weight most heavily, since they reflect how well a model performs when given a goal and left to execute rather than simply answering a single prompt. A jump in that category signals that Ling 3.1 Flash is targeting practical deployment.
Ling 3.1 Flash is an open-weight release, which means the weights are publicly accessible. That matters for adoption: developers can run, fine-tune, and audit the model directly, lowering the barrier for teams that need control over their inference stack or cannot send data to a third-party endpoint.