If you build FiveM maps in 2026, you have two real options for getting a custom MLO onto a server: the in-game builder approach—led by the newly-released SB Studio (StarBoys, April 2026)—or the Blender-plus-Sollumz pipeline that's been the de facto default for years.

SB Studio made noise at launch because it promises something genuinely new: building interiors inside FiveM itself, with no Blender, no XML round-trips, no server restart. Sollumz, meanwhile, just shipped 2.8.3 in March with native binary support for both Legacy (Gen8) and Enhanced (Gen9) assets. Both tools are real, both are improving, and neither is universally better. They optimise for different things, and the right pick depends on what you actually need to ship.

This post breaks down the trade-offs honestly so you can make the call yourself. If you want a deeper look at what's actually changed in Sollumz over the last 12 months, I covered that separately in Sollumz in 2026: What's Changed.

What in-game builders give you

The reference implementation right now is SB Studio (StarBoys, released April 10, 2026 on cfx.re). It's a paid, Tebex-listed, FiveM Asset Escrow resource—code partly visible, config only—that runs as a standalone resource on Game Build 3258 or above. There are other tools in this space (KUZ Shell Creator was the early reference for the shell-first workflow), but SB Studio is the one pushing the MLO-from-scratch direction.

The honest pros

  • Zero external software. If you already admin a server, you're already inside the tool. There's no Blender install, no add-on, no PyMateria prompt, no CodeWalker round-trip.
  • Live preview with real lighting and props. What you see is what streams. You're placing vanilla props in a real GTA V scene, not a Blender approximation.
  • Vanilla YDR / YBN finder and editor. Scan a zone, identify the props and collisions, load them, modify them in place. Face editing, normal flipping, real-time collision control. For "I just want to move this vanilla wall" workflows, this is unmatched.
  • Built-in texture library. 101 seamless textures ship with the script, loadable as a YTD. You can add your own. Skip the UV-and-material setup that takes 30 minutes in Blender for a simple prop.
  • One-click export to YDR, YBN, YTYP, YMAP, YMF. Live streaming, no server restart. KonflictStudio on the cfx.re release thread called it exactly what it is: "the barrier of entry has been pushed down."
  • Performance. The dev claims 0.00–0.01ms idle cost, with interiors spawned on demand only when actively editing.

The honest cons

  • The dev's own disclaimer flags a ~30% crash chance during the 30-second MLO build/load phase on a busy live server. Once the MLO is loaded, it's stable. But that deploy window is a real operational risk you should price in for production servers.
  • Asset Escrow, not open source. You can configure it; you can't fork it, audit it, or extend its code. If StarBoys disappears, the tool is gone.
  • Game Build 3258+ required. Older server builds are out. If your server is on a pinned build, this is a non-starter.
  • LOD 200, not global render. Fine for interiors, less flexible for very large exterior scenes. You're still subject to the 65,500 building pool cap (the dev clarified that the cap isn't hit because assets load as you approach, but it's worth knowing).
  • Geometry is limited to what the in-game tools can express. Grid editor, addon meshes, real-time UV control—but not freeform Blender-style modelling. If you want a curved organic interior or a hand-modelled vehicle, the tool fights you.
  • Tied to FiveM only. None of it is portable to single-player GTA V. The outputs are FiveM resources; you can't repurpose them on a non-FiveM storefront.
  • ~5,000 lines of code for a tool this ambitious. You're betting on one team continuing to maintain it. The 30% deploy crash rate is two months old and still flagged as an ongoing issue.

What the Blender / Sollumz workflow gives you

Sollumz is free, open source, on a documented release cadence, and just shipped 2.8.3 in March 2026. It runs as a Blender add-on, currently supporting Blender 4.0 through 5.0+. If you want a detailed breakdown of the recent release history, read Sollumz in 2026: What's Changed—for the purposes of this comparison, the short version is: it just got significantly better.

The honest pros

  • Full modelling power. Anything you can build in Blender, you can put into GTA V. Buildings, vehicles, props, character cloth (YLD), even complex multi-room MLOs with custom architecture. No geometry constraints.
  • Native binary import/export since 2.8.0 via PyMateria. On Windows, you can import and export .ydr, .ydd, .ybn, .ytyp, .yft, and .yld directly from Blender. The same .blend file can target both Legacy (Gen8) and Enhanced (Gen9) builds at export time.
  • Mature vehicle pipeline. Automatic shattermap generation on export (2.8.0+), VEHICLE / VEHGLASS shader handling, full collision and rigging control. The "Preview Windows" button in 2.8.0+ lets you verify shattermaps without going in-game.
  • Full collision, portal, room, and timecycle control. The 2.7.x series added two-way selection sync between Blender objects and YTYP entities, plus gizmos for MLO timecycle modifiers. MLO authoring is much more interactive than it was a year ago.
  • Vertex paint tools, name tables, gizmo-based editing. The 2.8.0 release brought a substantial vertex paint upgrade: channel isolation, multi-object painting, channel transfer, palette picker for tint shaders, gradient tool, and a new pie menu.
  • Portable outputs. Your .blend file and its exports work for both FiveM and single-player GTA V via CodeWalker. You retain full IP control of the source files—if you ever want to release on a non-FiveM storefront, that's not a re-architecture job.
  • Active development. Sollumz 2.8.3 was March 2026. 2.7.2 in June 2025. 2.7.0 in May 2025. The release cadence is steady, the GitHub issues are answered, and the project has a healthy contributor base.

The honest cons

  • The prerequisite is Blender. That's not a footnote—it's a six-month skill ramp for most people coming in cold. Sollumz is a Blender add-on; you can't use it without knowing the host application.
  • The 2.7-to-2.8 jump broke a lot of older tutorials. The single biggest community complaint in 2026 (per the FiveM Developers Facebook group and the cfx.re thread on 2.8.0) is that "half the issues people hit aren't Blender problems—they're missing one or two steps that 2.7 tutorials never covered."
  • PyMateria is Windows-only. macOS and Linux users fall back to the CodeWalker XML round-trip, which is exactly the workflow 2.8.0 was supposed to retire. If you're on a Mac, the native path is not available to you yet.
  • Asset streaming on a live server is still the most common source of "why won't it appear" frustration. You own every step—model, UVs, materials, collisions, exports, server folder, YMAP—and each one can fail in isolation. The diagnostic surface is fragmented across GitHub issues, the Sollumz docs, Discord threads, and scattered YouTube videos.
  • Diagnostic surface is scattered. When something breaks, you don't have a single "support" channel. You're reading the GitHub issues, the Sollumz docs site, and the Discord. For self-taught devs this is fine; for teams expecting a vendor relationship, it's not.

The three deciding factors

Neither tool is universally better. Here's the framework I use when someone asks me which one to use.

1. Volume and complexity of what you ship

Simple interiors—shops, apartments, small prop counts—are exactly what SB Studio's grid editor and texture library are tuned for. You'll get there in a fraction of the time compared to Blender. Vehicle conversions, multi-room MLOs with custom architecture, or anything requiring freeform modelling: only Sollumz gives you that surface. SB Studio's grid editor and addon mesh tools impose real structural limits.

As a rule of thumb: if you can describe the interior as "boxes and props," in-game builders win. If you need curves, custom architecture, or organic shapes, you need Blender.

2. Where your team sits on the skill curve

If your team is server admins with no 3D background, SB Studio removes the Blender dependency entirely. That's a meaningful productivity unlock—a small team can now ship interiors that previously required either commissioning them or hiring a 3D artist.

If you already have a Blender-capable artist on the team, Sollumz gives them a full DCC and they won't feel constrained. The cost of Blender proficiency is a sunk cost at that point.

3. Whether you need FiveM-only or cross-target

SB Studio outputs are tied to FiveM and the Asset Escrow system. You can't repurpose them for single-player GTA V, and you don't own the tool's code. Sollumz outputs are portable, you retain IP of the source .blend files, and the same MLO can be sold on a non-FiveM storefront or used in a single-player context.

If you ever want to release the same MLO on a non-FiveM marketplace, or if you want to keep ownership of every asset, that distinction matters. If you're purely building for your own server and you don't intend to sell the interiors, it's less important.

A quick decision matrix

If you can answer "yes" to most of these, the in-game builder is the right tool for that project:

  • The interior is a small to mid-size room with mostly rectangular geometry.
  • You're using vanilla props or simple addon meshes—no hand-modelled unique architecture.
  • Your team has no Blender experience and no time to learn it.
  • You're not planning to sell the MLO outside FiveM.
  • You're on Game Build 3258 or higher.

If you can answer "yes" to most of these, the Sollumz / Blender pipeline is the right call:

  • You need custom architecture, organic shapes, or detailed modelling.
  • You're doing vehicle conversions, complex multi-room MLOs, or character cloth (YLD).
  • You (or your team) are already comfortable in Blender.
  • You want full IP ownership of the source files.
  • You plan to sell the MLO on a non-FiveM storefront or use it in single-player.
  • You're on Windows and can use the native binary path.

Conclusion

Both tools are real and both are improving. SB Studio is two months old—the 30% deploy crash rate is a known, disclosed issue, and the "in-game, no Blender" promise is genuinely new for the scene. The release video got it right: this is the moment when custom interiors stop being the exclusive domain of 3D artists. That's a real shift.

Sollumz is mature, but its learning curve hasn't shrunk, and the gap between "tutorials that exist" and "tutorials that match the current version" is still wide. The good news is that 2.8.x is the best version the tool has ever shipped—if you're going to learn it, learn it now.

Don't pick based on hype. Pick based on what you actually need to ship and what your team can operate. Both approaches are valid; neither makes the other obsolete.

If you decide the Blender route is right for your project, I've put together a full walkthrough of the Sollumz 2.8.x workflow for FiveM MLOs—from first import through to a streamed resource on a live server. The guide is here.