Skip to content
GTA666
intermediate9 min readUpdated 2026-09-09

Picking a Framework: QBCore, ESX or ox

QBCore vs ESX vs ox_core vs standalone — the trade-offs, the script ecosystems, and how to choose the one you can still live with in a year.

Quick answer

A framework decides which of the community's tens of thousands of scripts you can install without rewriting, so it is the least reversible choice you will make. ESX is the oldest and broadest, QBCore is the modern roleplay default with the largest current script market, and ox_core is the cleanest and most demanding. Standalone is viable only if you have a developer.

This is the choice that constrains every later one. Not because the frameworks differ enormously in what they do — they all give you characters, money, inventories and jobs — but because the community’s scripts are written against a framework. Pick QBCore and a QBCore mechanic job installs in five minutes. Pick something else and that same script is a porting project.

As throughout this track, this is the FiveM/GTA V ecosystem. GTA 6 private-server support has not been announced, and no framework exists for it. If a platform ever appears, expect this exact pattern — a couple of dominant community frameworks and an ecosystem locked to them — to repeat.

What a framework actually provides

Under the surface, all of them do roughly the same job, described in how servers work:

  • Identity and characters — mapping a player’s identifiers to one or more characters.
  • Money — cash, bank, and often dirty money, persisted to MySQL.
  • Inventory — items, weight, containers, and the shared item definitions every script references.
  • Jobs and grades — the permission structure most roleplay content hangs off.
  • A function and event API — what third-party scripts call.

The differences that matter are ecosystem size, code quality, and how opinionated the design is.

ESX

The oldest of the big frameworks, and the reason a lot of the conventions exist at all. ESX’s documentation covers es_extended, its client/server/shared structure, its events and its player data model, and the project describes itself as free and open source with a very large installed base.

Strengths: an enormous back catalogue. If a roleplay concept existed before 2021, there is an ESX script for it. Endless tutorials. Easy to hire for.

Weaknesses: age. Years of layered conventions, a lot of old scripts written to standards nobody would accept now, and a long tail of tutorials referring to versions that no longer exist. Legacy ESX is actively maintained, but you will meet code that predates the maintenance.

Choose it if: you are building something conventional, want the widest possible pool of existing scripts and guides, and are comfortable curating aggressively.

QBCore

The framework most new roleplay servers start on today. Its docs live at qbcore.org/docs, and its core is open on GitHub as qb-core.

Strengths: it was designed for roleplay specifically, so the defaults suit an RP server rather than a generic gamemode. The current script market is overwhelmingly QBCore-first — most paid resources ship a QBCore version and maybe an ESX one. Its modular qb- resources are readable and easy to fork.

Weaknesses: the ecosystem’s quality is wildly uneven, and a stock QBCore server is close to being every other stock QBCore server. Distinguishing yours takes real work — see launch day and growth.

Choose it if: you want the biggest supply of ready-made roleplay content and the shortest route to a working server. For most first servers, this is the sensible answer.

ox_core

From the Overextended project, whose docs at overextended.dev cover ox_core, ox_lib, ox_inventory and oxmysql.

Strengths: by a distance the best-engineered option. Careful design, sane defaults, server-authoritative habits, and a real attitude to performance and security — which matters more than it sounds, as performance and security explains. ox_inventory in particular is used widely well outside ox_core servers.

Weaknesses: the smallest ecosystem of the three. You will convert scripts or write them. It is also more opinionated: it will not let you do things the sloppy way, which is a virtue on a good day and friction on a deadline.

Choose it if: you have a competent developer, you intend to build custom content rather than assemble it, and you plan to run this server for years.

Standalone

No framework: just the base resources and whatever you write. Perfectly valid, and how a few of the most distinctive servers work.

Strengths: no bloat, no inherited assumptions, no framework upgrade breaking forty resources at once. Genuinely fast.

Weaknesses: you build character persistence, money, inventory and jobs yourself, and then nothing off the shelf fits. This is a development project, not a server setup.

Choose it if: your concept is narrow — a racing server, a survival server, a minigame server — or you have a team that would rather write than integrate.

The libraries everyone runs anyway

Regardless of framework, two Overextended resources have become near-universal: oxmysql, the database connector most modern scripts expect, and ox_lib, a shared utility and UI library. Many resources list them as dependencies outright. Install them early, and read scripts, MLOs and vehicles before you install anything else.

How to actually decide

Do this rather than reading opinions:

  1. List the ten scripts your server concept depends on. The specific ones — the inventory, the phone, the police MDT, the dispatch, the housing.
  2. Check which framework each supports. This usually decides it in five minutes, and it decides it on evidence rather than taste.
  3. Check the licensing and support of anything paid, and whether it is escrowed.
  4. Install the framework via txAdmin’s recipe deployer, which will build a working server from a template in one flow, per the txAdmin guide.
  5. Play it for a week before committing. Not test it — play it, with five friends, for real.

Checklist

  • Do the scripts I need exist for this framework today?
  • Is it actively maintained, with commits in the last month?
  • Can I read its core code and roughly follow it?
  • Is there somewhere I can ask questions and get answers?
  • If my developer disappears tomorrow, could I hire a replacement who knows it?

That last question is the one people skip and later regret.

Takeaways
  • Pick for the script ecosystem you need, not the code you admire
  • QBCore has the largest current market of roleplay scripts; ESX has the deepest back catalogue
  • ox_core is the best-engineered option and the smallest ecosystem
  • Migrating frameworks later means rewriting or rebuying most of your server
  • Whatever you choose, ox_lib and oxmysql are near-universal companions

Frequently asked

Can I switch framework later?

Technically yes, practically it is a rebuild. Your database schema, every script you bought or wrote, and your staff's muscle memory are all framework-shaped. Servers that migrate usually treat it as a relaunch with a wipe, and lose players doing it.

Is standalone a real option?

Yes, if you have a developer and a narrow concept. Standalone servers are faster, cleaner and completely yours — and you will write character persistence, inventory and jobs from scratch, then find that no community script fits. Choose it deliberately, not by accident.

What are ox_lib and oxmysql?

Shared libraries from the Overextended project. oxmysql is the database connector most modern resources expect, and ox_lib provides common UI and utility functions. You will end up running both regardless of which framework you pick.

Sources
  1. ESX Framework documentation · ESX Framework
  2. QBCore documentation · QBCore Framework
  3. Overextended documentation · Overextended
  4. Setting up a server using txAdmin · Cfx.re
Day-one whitelist

Get recruited when GTA 6 private servers open.