Prime Intellect Releases Prime Agent: An Open-Source RLM Harness Where Sub-Agents Are Function Calls Inside Persistent IPython Kernel

Prime Intellect has released Prime Agent, an open-source Recursive Language Model (RLM) harness that uses a persistent IPython kernel for sub-agent function calls, achieving a 95.5% score on ARC-AGI-3, surpassing human expert baseline. It supports multiple AI APIs and self-hosted models, suitable for mid-size to large engineering teams for complex coding tasks.