Getting started with Emergent

This page helps you get started with Emergent (emergent.sh) step by step: from your first account to a published app on your own domain. Also read the complete Emergent guide for background, and prompting in Emergent for how to brief the agent effectively.

Creating an account and free credits

You start with a free account on emergent.sh. According to the pricing page, the Free plan gives you 10 free monthly credits, with access to the platform's core features, support for web and mobile apps, and direct access to the latest models. There's no GitHub integration on Free — that's only available on the paid Standard plan (see Emergent pricing).

In practice this means: Free is fine for discovering how the agent works and whether this way of building suits you, but with 10 credits a month you can't iterate endlessly. Decide in advance what you want to test, so you use your credits purposefully.

Formulating your first task

Unlike a chatbot where you ask small, separate questions one at a time, Emergent works best with a clear, well-defined task given all at once: what should the app do, who is it for, and what counts as a good result? Because the agent gets to work independently — writing code, testing, deploying — a vague first prompt is one of the costliest mistakes you can make: the agent keeps building on assumptions you never checked.

A good first task includes at minimum:

• The purpose of the application in one or two sentences.
• Who the users are and what they need to be able to do.
• What data is recorded or displayed.
• What counts as a successful result (acceptance criteria).

See prompting in Emergent for a detailed structure and examples of strong versus weak briefs.

Following the agent while it works

Because Emergent positions itself as an autonomous agent (see the "E3" announcement on its own blog, 8 June 2026), it generally completes a task more independently than a tool where you approve every step. Still, keep an eye on it:

• Follow the activity log or progress view in the interface, so you can see what steps the agent is taking.
• Step in as soon as you notice the agent making an incorrect assumption (for example, an incorrect data model), rather than waiting until the whole task is finished.
• Use Emmy, the in-product assistant Emergent introduced in July 2026, for help writing follow-up prompts, troubleshooting and guidance during the build process.

Testing what the agent delivers

After a task is finished, don't blindly assume "it works because the agent says it works." Walk through the application yourself:

• Manually test the key user flows, including error scenarios (incorrectly filled-in fields, no internet connection, empty datasets).
• Check authorisation: can user A access user B's data that they shouldn't have access to?
• Check that data is stored and displayed correctly, especially in forms and payment flows.
• If you're unsure about the underlying technology or data security, ask Emergent directly — the exact technical stack and the credit consumption model are not fully documented on the public pages we could consult.

Publishing and connecting a custom domain

According to a blog post from Emergent ("How to Buy a Custom Domain on Emergent?", 13 May 2026), you can purchase a custom domain and link it to your published app. In practice, publish first to Emergent's default (sub)environment for internal testing, and only connect a custom domain once you're confident the app is ready for visitors.

GitHub integration

GitHub integration and "Fork tasks" are listed on Emergent's pricing page as features of the paid Standard plan and above — you don't have access to these on the free plan. What "Fork tasks" precisely entails is not further explained on the pages we could review; check with Emergent if version control and code ownership are critical to your project. Once you're on a paid plan, enabling GitHub sync is advisable as soon as you're building more than a small experiment: it gives you a version history and a fallback option if a change goes wrong.

Checklist before going live

Go through these points before rolling out an Emergent app broadly:

• All key user flows have been manually tested, including error scenarios.
• Authorisation has been checked: users can only access their own data.
• You know who will maintain and further develop the app once it's live.
• You've asked Emergent how ownership, export and backup of your project are arranged.
• If you process personal data of EU citizens: you've requested GDPR and hosting information from Emergent — this is not publicly documented on the pages we could consult (see the Emergent guide for more detail).
• You've checked your plan's credit consumption against your expected usage intensity, so the app doesn't hit a limit mid-peak.
• A custom domain is connected and tested, including SSL/HTTPS.
• If in doubt about security or scalability, you've had a technical review carried out by someone with development experience.

Not sure whether your project is ready for production, or want an independent check before going live? Brandable helps organisations in the Brainport region with this kind of project, even when the foundation isn't built in Lovable but in another platform such as Emergent — see web design or book a consultation.

Want to know more about how to effectively brief the agent? Continue to prompting in Emergent, or first read the broader background in the Emergent guide. Considering using Emergent commercially? Take a look at Emergent for businesses.