Glossary · Architecture

Browser execution layer

The infrastructure that turns an agent’s goal into deterministic, replayable browser action — the layer between your LLM and a real browser.

What is browser execution layer?

A browser execution layer is the system that sits between a language-model agent and a live browser, taking a natural-language goal and producing reliable, repeatable action against real pages. Twin’s execution layer bundles the DOM compiler, planner, skill cache, deterministic replay, credential vault, human handoff, and session video behind a single token-efficient API.

Why it matters

Without an execution layer, every team rebuilds the same brittle glue between an LLM and a browser and pays the model on every run. Consolidating that into one layer — and adding a semantic cache on top — is what turns ad-hoc agent scripts into production automation whose cost falls with scale.

See it in context: read how Twin compiles and replays a run, follow the cost-cutting guide, or browse use cases and comparisons.

Run your first skill

Give an LLM agent a real browser, compile the workflow once, and watch the marginal cost fall as the cache takes over.