How to Use AI Agent?

Although EMILY.RPA provides many No-Code + Low-Code automation modules that can handle the vast majority of situations, when facing dynamically changing automation targets — such as frequently redesigned websites or mobile-captured documents — you'll need AI Agent modules that can think and act in real time during execution.

Since AI Agents are powered by large language models (LLMs), there are GPU computation costs involved. Depending on the model chosen, the per-token cost for input and output varies. Model selection relates to the complexity of the target — typically, more complex targets require more advanced models. You can try different brands and versions of models during the design phase.

EMILY.RPA allows users to drive Agents using their own API Keys, or to purchase credits through EMILY.RPA to offset the cost of each execution. By purchasing credits through EMILY.RPA, you gain access to enterprise-grade API Keys with guaranteed Zero Data Retention from model providers and priority access. You can seamlessly switch between different model providers, with balances automatically deducted based on each model's listed price — no credit card binding required.