mrkeyoor.com_
Sun 25 Oct 18:52 UTC
Dev Toolsevaluationupdated 25 Oct 2026

armorpaint review

ArmorPaint is a standalone application for physically based texture painting on 3D models. Its manual documents brushes, layers, node materials, baking, mesh adjustments and texture export presets. This is a documentation-based review: our recorded lab run did not install, build, test or launch the application.

+29stars / 7d
Verdict

ArmorPaint is worth evaluating for dedicated PBR painting and configurable texture exports if you can check the workflow on your own hardware. The README distinguishes paid distributed binaries from a developer-focused Git checkout that may be unstable. Our October 16, 2026 lab record for d91edb4 contains no execution steps, so it establishes neither setup success nor product failure. Start with one representative asset and verify import, painting, save/reopen and engine export before committing a production project.

We ran it

Screenshot of armorpaint (armorpaint.org)

Answers from our run

Did you run armorpaint yourself?

No. Its code is C, and it carries no manifest our lab installs from, and no Dockerfile, so there was nothing standard to install, build or test. This review is written from the repository's own documentation.

Who should not use armorpaint?

Users expecting free official distributed binaries: the README says those builds are paid to help fund development.

What are the alternatives to armorpaint?

Blender, Material Maker. ArmorPaint is worth evaluating for dedicated PBR painting and configurable texture exports if you can check the workflow on your own hardware.

Setup3/5The manual documents portable desktop binaries and the README provides platform build routes. Our October 16, 2026 record for d91edb4 contains no execution steps, so setup ease remains unverified.
Docs4/5The supplied manual gives specific import/export formats, channel-packing presets, GPU guidance and local-versus-cloud processing settings. Our lab did not validate those workflows.
Community2/5The supplied issues show maintainer responses and follow-up, including a reported grid fix. They do not establish adoption, response-time consistency or representative support quality.
Maturity3/5The release page identifies 1.0 at tag 26.09, while the manual labels mobile experimental, sculpting work in progress and live-link plugins in development. Issue 2052 was reopened on October 24, 2026.

Who it’s for

3D artists evaluating a dedicated PBR painter for Windows, Linux or Apple silicon macOS.
Technical artists who need configurable texture channel packing for an engine pipeline.
Developers comfortable building a potentially unstable Git checkout with a platform compiler.
Artists considering paid desktop binaries who can validate their GPU, imports and exports before adopting the tool.

Who it’s NOT for

Users expecting free official distributed binaries: the README says those builds are paid to help fund development.
Teams requiring the development checkout to come with a stability assurance: the README explicitly warns that it may not be stable.
CPU-only painting workflows: the manual says painting runs on the GPU.
Artists whose main requirement is finished sculpting: the manual labels that workflow work in progress.
Teams requiring proven mobile or pixel-perfect behavior without their own evaluation. Mobile builds remain experimental, and the pixel-grid fix does not resolve every brush-rasterization concern.

Setup reality

In our recorded run on October 16, 2026 at 02:49:49.494 UTC, the lab inspected commit d91edb4 (full SHA d91edb475ed5ffe61cde52d07298aa843bd9b0b5). The harness recorded "not_runnable" because it had no supported ecosystem for this C project and detected no Dockerfile. The raw record contains no execution steps. No installation, dependency download, build, tests, dependency audit, installed-binary check or painting workload was recorded. This is a harness limitation, not evidence that ArmorPaint cannot build or run.

The README requires Git and a platform toolchain: Visual Studio with Clang tools on Windows, Clang plus dependencies on Linux, Xcode on macOS or iOS, and Android Studio on Android. The supplied extract does not enumerate Linux system packages. For desktop binaries, the manual describes unpacking and running a portable application; we did not verify that path.

The manual lists Intel HD4000 as the minimum for 4K painting and recommends an Nvidia RTX 2080 with 8 GB or better for 16K painting. These are the project's requirements, not our performance measurements.

Choose the distribution route before judging setup

ArmorPaint has two different setup paths. Its README says distributed binaries are paid to help fund development, while the Git repository is aimed at developers and may not be stable. The manual describes desktop binaries as portable: unpack the archive and run the application. Building the checkout requires Git and the appropriate platform compiler.

That distinction matters for a 3D artist deciding whether to evaluate ArmorPaint. A source-build review cannot establish how smoothly the packaged application runs, and our lab did not execute either route.

What our recorded run establishes

Our recorded run inspected commit d91edb4 (d91edb475ed5ffe61cde52d07298aa843bd9b0b5) on October 16, 2026 at 02:49:49.494 UTC. The commit timestamp in the record is October 15, 2026 at 23:20:31 +02:00.

The harness marked the project not runnable because it lacked a supported ecosystem for the C project and detected no Dockerfile. The raw execution-step object is empty. There is no recorded install command, dependency download, build, test execution, dependency audit, installed-binary check or measured painting/export workload. The sandbox had 3 CPUs and 8 GB of RAM; those resources do not establish GPU capability or application performance.

This result limits our review. It does not demonstrate a broken build, failed tests, missing upstream tests or a security finding. The current README and manual extracts describe the project, but they are not a verified snapshot of that exact commit: retrieval of its commit-specific README failed.

The documented painting and export workflow

The manual describes node materials and brushes, paint and fill layers, masks, decals, clone and blur tools, and painting in a 2D texture view. Projects save meshes, layers, materials and brushes in an .arm file; the Pack Assets option includes external assets such as textures.

Documented mesh imports include OBJ, FBX, Blend, STL, glTF and GLB. OBJ import can split by UDIM tile. The mesh tools include UV unwrapping and basic modifiers. If you change UVs inside ArmorPaint, the manual says to export the modified mesh alongside the textures so they map correctly.

Texture exports support 8-bit PNG or JPEG and 16-bit or 32-bit EXR. The Unreal preset packs occlusion, roughness and metallic; the Unity preset packs metallic, occlusion and smoothness. Custom presets configure individual RGBA channels and can be imported from JSON. These documented controls make channel packing a useful part of an evaluation, but we have not checked the resulting files in an engine.

Baking includes ambient occlusion, curvature, normal and height maps, among other modes. Several modes require ray-tracing hardware. The manual also warns that its path-traced viewport favors performance over correctness. Treat that viewport as a documented preview option rather than a validated final-render reference.

Check GPU requirements and where processing happens

Painting runs on the GPU according to the manual. Its 4K minimum is Intel HD4000; its 16K recommendation is an Nvidia RTX 2080 with 8 GB or better. These figures are vendor guidance, not measured throughput or quality results from our lab.

The manual says neural nodes process locally and can operate without an internet connection after model downloads. It recommends at least 6 GB of video memory, while individual listed models range from Real-ESRGAN at 1 GB to Qwen at 16 GB. Hunyuan3D is listed as Windows-only with 12 GB required.

The console has a separate choice that affects this distinction. The manual describes local Qwen processing, but selecting claude, grok or codex sends console processing to the cloud and requires access configured on the computer. Do not assume every AI-assisted operation stays local merely because neural nodes do.

Read issue histories through their latest supplied updates

Issue 2111 no longer supports an unqualified claim that grid snapping remains broken. On September 15, 2026, the maintainer reported a fix at fe6b787 and closed the issue as completed. The same comment says the brush-rasterization concern still needs work and suggests the decal tool for a square brush. We did not test that fix or establish which packaged builds include it.

Issue 2031 was closed as completed on September 2, 2026 after follow-up by email. A September 8 comment from the reporter still describes Apple Pencil trouble on iOS 18. This is a version-specific report, not evidence that OBJ import or painting fails on every current iOS device. The manual continues to label iOS and Android builds experimental.

Issue 2052 remains open in the supplied extract after being closed and reopened on October 24, 2026. The original report concerns Blender project imports on Linux Mint 22.3, ArmorPaint v1.0 alpha 2026/04/11 at 8f273f7 and Blender 5.1.0. Later comments discuss temporary-file paths. These reports justify checking your own Blend import route; they do not establish a universal import failure.

The release page identifies version 1.0 with tag 26.09. That label alone does not verify workflow reliability. Sculpting remains marked work in progress in the manual, and live-link plugins for Unreal, Unity and Blender are described as in development.

How to use the ratings

These are editorial evidence assessments, not benchmark results. The scale is 1 for substantial documented barriers, 2 for limited support or unresolved evaluation gaps, 3 for a documented usable route with qualifications, 4 for detailed support across the reviewed needs, and 5 for comprehensive support corroborated by relevant execution evidence. Each category applies that scale to its own evidence; unknown adoption or compatibility earns no credit.

Setup receives 3/5 because the documentation provides a portable binary route and platform build instructions, although neither was executed here. Documentation receives 4/5 for its concrete workflow, format and hardware guidance. Community receives 2/5 because the supplied issue discussions show maintainer responses but do not establish adoption or representative support quality. Maturity receives 3/5 because the 1.0 release coexists with explicitly unfinished features and a reopened import report.

For a practical next step, use one asset representative of your work. Check its import path, required brushes and layers, save/reopen behavior, texture channel packing and final appearance in your target engine. If local processing matters, check the selected console model as well as neural-node settings.

Alternatives

ProjectWhat it isPick it when
Blender gh↗A free and open source 3D creation suite covering modeling, rigging, animation, simulation, rendering and other pipeline stages.Choose Blender when the work also requires modeling, animation or rendering in a broader 3D suite. The supplied source does not establish a texture-painting feature comparison.
Material MakerA Godot-based tool for procedural texture creation and painting 3D models, with textures and brushes described through interconnected nodes.Choose Material Maker when procedural texture creation and a node-based texture-and-brush workflow are central to the evaluation.

What people are saying

  1. [github-trending] armory3d/armorpaint

Sources

  1. ArmorPaint README
  2. ArmorPaint manual
  3. ArmorPaint 1.0 release
  4. ArmorPaint license
  5. Pixel-perfect drawing issue 2111
  6. Blender project import issue 2052
  7. iOS OBJ import issue 2031

More dev tools reviews

thinkrail · firebase-tools · brotli · agentskills · hallmark · ds-harness-remote · the whole board →