The initiative targets teams working on scientific validation, hypothesis testing, and technical development. Beyond the compute credits, selected partners receive dedicated engineering support with a three-hour service level agreement and the opportunity to showcase findings within the Apodex research community. The program aims to move beyond standard generative AI by providing tools that perform deep literature synthesis and complex analytical reasoning.

At the center of this support is Apodex-1.0-H, a model built to coordinate up to 150 sub-agents over 15,000 steps. Founder and CEO Tianqiao Chen stated that the system prioritizes verification to reduce unsupported conclusions, separating the solving process from the auditing of evidence. By allowing specialized agents to cross-check findings before delivering a final result, the company intends to provide a more reliable framework for frontier science tasks where accuracy is paramount.