Apodex 1.1: AI Model Challenges Larger Rivals with 150-Task Capability

A relatively small AI lab has made significant strides in AI development with its release of Apodex 1.1, a model capable of juggling up to 150 tasks at once while running on hardware that users may already own.

The Apodex team positions executable environments as the primary scaling surface for long-running AI agents and claims their model's benchmark scores put it in direct conversation with models from labs that have considerably more resources.

The model, detailed in an arXiv paper (2608.23283), posted benchmark scores of 38.5 on APEX-Agents and 78.8 on GDPVal, two benchmarks designed to measure agentic task completion and general reasoning under realistic conditions.

The Apodex team also released open weights for the Mini variant of their model, which runs on 35 billion parameters, a fraction of the parameter counts associated with frontier models from major labs.