Intel / article

Unreal Agent

Source: unreallabs.ai
Author: trollied
Date: Published Sep 22, 2026

Context

Unreal Labs built Unreal so an AI coding assistant can keep working while its tools run. When the agent issues a tool call, the surrounding software immediately logs it as in progress and runs it in the background. The software, not the model, handles waiting and polling. When the tool finishes, the result is added to the session log and the model is called again. Meanwhile the user can steer the agent or start long work, such as setting up a development environment, while the agent explores the codebase. The company attributes up to 40% lower cost than Codex to a minimal with compact tool results and more tool work per model turn. Those figures come from one model configuration, mostly on coding tasks, so they are the company's account rather than a general result.

Terms in this piece · Glossary

AI agent — An AI system that doesn't just answer once but works toward a goal in a loop — taking actions, reading the results, and deciding what to do next.
agent harness — The scaffolding around a model that turns it into a working agent — the loop, the tools it can call, and the rules for when to stop.

Read the source unreallabs.ai