What is WebAgent

WebAgent is a general-purpose agent platform for the browser — an API that lets programs complete tasks in a real browser, from a single atomic search to a goal that runs for a month.

The four APIs

WebAgent exposes 4 peer APIs; pick whichever matches what you want to do:

• DoAnything — give one natural-language instruction; the agent picks its own tools and path and completes the task in a browser. The artifact shape is open-ended.
• DeepResearch — give a research topic; it runs multiple retrieval rounds and cross-checks, and produces a cited, confidence-scored report (final.md + citations + confidence).
• WebSearch — give a batch of queries; it fetches across engines, deduplicates, reranks, and returns structured search results (optional summary).
• Track — give a monitoring intent; it re-fetches on a schedule, compares against a baseline, and notifies you on change (a snapshot stream + change notifications).

Shared capabilities:

• Profiles — reusable login state across sessions. No re-logging in every run.
• Workspaces — a persistent file system the agent can read and write to.
• Schedules — cron, interval, event-triggered, or autonomous (the agent decides when next to run).
• SSE event stream — the same run.* events that drive the Console, streamed directly to your code.

What it is not

• Not a low-code automation builder. There is no canvas. You wire up tasks in code (or via the Console as a prototyping aid).
• Not a hosted LLM API. Bring your task; WebAgent picks an LLM and pays the bill on a credits model.

Three product surfaces: Console / OpenAPI / SDK

All APIs are exposed through the same three surfaces, with 1:1 capability parity and a shared resource layer / event stream / billing:

You can use …	… to do
The REST API (OpenAPI)	Anything. Console and SDKs are just clients. api.eak.eazo.ai/v1/... + Authorization: Bearer wa_...
Python or TypeScript SDK	Same surface, idiomatic types, retries, streaming, wait_for_done.
The Console	Prototype tasks visually; non-developers welcome; Get Code dialog hands you working snippets.

API-developer-first — the product is the API. Console is a convenience layer, not a separate product surface; no Console-only privileged endpoints.

Resource model

The 4 APIs share one event stream and billing, but each has its own resource model. DoAnything uses session + run:

Session                            (one container; holds a browser, profile, workspace)
└── Run #1  status: completed      (one instruction; lifecycle has 7 states)
└── Run #2  status: running        (a follow-up instruction in the same session)
    └── events: SSE stream         (status_changed, message, action.*, screenshot, …)

A session owns the runtime resources (browser, profile, workspace). Each run is one instruction; you can submit follow-up runs against the same session and they share state. The run lifecycle has seven states (pending, running, awaiting_input, paused, done, failed, canceled); see DoAnything.

DeepResearch / WebSearch / Track differ: DeepResearch / WebSearch are standalone runs (one-shot artifact, no session); Track uses long-lived monitors. Each feature page covers its own.

Next steps

• Quickstart — 5 minutes from sign-up to first SSE event.
• Authentication & API keys — how wa_ keys work and how to scope them.
• DoAnything — the core resource model.