Website and Patreon access

Throughout the site you will see the model Generation mentioned. This is the general set of features of a model, and in general a bigger generation number means a model with more recent updates.

You can find a general changelog of the Generations at the dedicated page.

All models in the Model Page have a developement/release status. The legend is the following:

  • In developement: the model is not finished yet, check the page to see the status in a couple of weeks.
  • In developement (Beta testing available): the model is not finished, but a Beta test is available for some people. I will send invitations to people myself (see FAQ What’s the Beta section? below). In about a couple of weeks the model should become available on Patreon.
  • Preview version on Patreon: a preview version is available on the Patreon archive, the public release might happen in a couple of weeks/some months, depending on how much bugs are found by patrons.
  • Released: the model is freely available. Patrons can still download these models from the Patreon archive, with uncapped download speed.
  • Patreon only: the model will be a Patreon exclusive, and available on the Patreon archive. This is usually the case for Original Characters or models that were in developement but were never finished.

In general, if the download button in the Model page is gray, it means it can not be downloaded. Check the page later to see its status, or contact me for more informations!

The new Patreon archive can now be found on the page:

You can access the link above with your Patreon credentials!
Just temember to click on Allow on the Patreon access page.

  • Unfortunately, due to the API limitations, you’ll need an active subscription and at least a 1€ pledge. Therefore be sure to currently pledge at least 1€ before trying to access the archive. Consider also the currency changes, as 1$ is not 1€ in general.
  • Moreover, for security reasons, while the payment is pending (i.e. at the start of the month or soon after pledging with some payment methods), the login might fail. If this happens, please wait a bit. When your payment won’t be in the “Pending” status anymore, you should be able to access.
  • When a month is paused you will get free access if and only if you did not cancel the subscription the month before. Unfortunately if you cancel the subscription before the end of the month, you will need to pledge again to get access to the archive.
  • If any other error appears, please clear your browser cache and logout/login again on Patreon, or try with another browser.
  • If you can’t solve the access problems, please contact me (Twitter and Discord will be faster for replies).

The alpha/beta testing is based on this section, which is password protected. Some invitations are sent to creators, but if you want to have the access please contact me, and I will consider it. At the moment I can accept only well established creators.

You can check if your account have access to the Beta page in your Profile page, in the About panel. If you have access, the account status should be Beta Tester. If you were invited, but the account status is not correct, please contact me.

Note that this do not substitute Patreon early access! It is more for specific features testing and mainly for creators. Beta and Patreon early-access might share the same models, but usually once the beta testing is done, the models will only be on Patreon (and eventually publicily released).

There will never be any advertisement on this webpage, and I disabled all possible trackers like Google Analytics. So enjoy a true-privacy experience (only Patreon for obvious reasons will know you were on the website if you want to access the archive).


If you manually delete an Outfit collection (Delete Hierarchy in the Outliner), some Orphan data (data that are not available on the scene but can still be found in the file, for instance from the Blender File mode in the Outliner) might still be around, affecting performance and increasing the size of your file. There are several ways to remove Outfits from the scene, using MustardUI, without leaving these Orphan data around.

  • Clean Model: if you plan to use only one Outfit, and you already selected Extras and Hair in the MustardUI, you can use the Clean Model tool to remove all other outfits. Check the Documentation for more informations about this tool.
  • Delete Outfits with right click: if you want to remove just some Outfit collections, enter configuration mode from the Settings panel of the MustardUI, and then right click on the collections and use the Delete Outfit button. Note that Remove Outfit just delete the outfit from the UI, but keeps the Objects in the scene.
  • Delete Outfits from the Outfits list: if you want to remove just some Outfit collections, enter configuration mode from the Settings panel of the MustardUI, and go to the Outfit section. There is a list of Outfits available, with two buttons. The X button will just remove the Outfit from the UI, but keeps the Objects in the scene; while the “trash icon” button completely deletes all Objects in the Outfit collection from the scene.

If you manually delete a Model from the Outliner, some Orphan data (data that are not available on the scene but can still be found in the file, for instance from the Blender File mode in the Outliner) might still be around, affecting performance and increasing the size of your file.

Thus, to completely remove a model I highly suggest to use the MustardUI Removal UI tool. Check the Documentation for more informations about this tool.

You can report the bugs on the #bug-report channel on the Discord server.


If the author of the model provides the mustard_ui.py script and register it, you do not have to install any addon. Usually this happens when you see that the model features MustardUI. In general, it is better not to install MustardUI as an addon, as it might clash with specific models UI updates if you don’t update the add-on every time.

If you know the model provide a MustardUI and for some reason the mustard_ui.py file is not available in the Blender Text Editor list of files, you can install the UI as an add-on, downloading the latest version from the releases page. However, we highly encourage you to contact the model creator to report the missing script bug.

Make sure to follow these steps:

  • after append, save the file and restart Blender. Now the UI will appear when you click on the armature.
  • if you want the UI to be available also when you’re not selecting the armature: click on the model armature (any of the model bone either in Object or Pose mode), and deselect the Viewport Model Selection in the MustardUI Settings.

Multiple characters using the MustardUI in the same scene are supported only for MustardUI versions > or equal to 0.20.
In MustardUI versions before 0.20, only one model UI will work, and you will have to change the properties of the second manually.

When you disable the morphs in the default settings, also shape keys are disabled to increase performance. You can change this behaviour disabling the Mute Shape Keys setting as in the image (see the Documentation, MustardUI and Tools -> Morphs for more details).

By default, when you disable the Morphs, all drivers are disabled to increase performance. You can choose to keep some drivers enabled by adding the string “MustardUINotDisable”.

For instance, if you have a shape key named “Hand” with a driver you made on it, just rename it “MustardUINotDisable Hand”, and the driver will work even if you disable the Morphs.

To generate a log file, enable Maintenance tools in the Settings panel of MustardUI. A button called “Create Log file” will appear, that can be sent to the model provider for assistance.

The log will contain informations on your CPU/GPU, and the current enabled/disabled features.

Note: NO personal informations, such as installation folders or system user names, is logged!

Sure you can! The Configuration panel is in the Settings (if Maintenance Tools is enabled). Just click on the UI Configuration button, and start modifying the UI. More informations are available on the GitHib Configuration guide.

Due to some changes in the Blender Python (for which I have no control on), the UI breaks when it renders any part of the UI that uses the HAIR icon. To solve this issue, update the UI to the latest version.
You can just open the Text Editor, find mustard_ui.py in the list of files above, select everything and paste the content of the new mustard_ui.py file you can find in the most recent MustardUI release.


Due to the high amount of morphs (and therefore Blender drivers) in the model, the model might be slow with default settings, especially if you plan to make an animation. The hints below will help you get decent performance during production!

  • Enable MustardUI Simplify (see Simplify in the MustardUI section).
    Alternatively, if Simplify is not available or you want to optimize the model manually, check these videos to help you tweak the performance of the model during production.
  • You can also enable, in the Render settings tab, Blender Simplify. Set Max Subdivisions to 0 (only for viewport). It is always better to disable Simplify before the final rendering.
  • Some add-ons are known to slow down the viewport when active, even if not actively used in the scene. Some of them are Wiggle bones and Spring bones. In general, it is always advised to disable all unused add-ons to reduce the probability of dealing with these issues.
  • A lot of outfits in the scene, even if disabled, can decrease viewport performance. Therefore, I advise you to delete the outfits you do not want to use in the current scene. If you want to preserve just one outfit, check the Clean Model function (check the Clean tool in the MustardUI – Settings section of this documentation).
  • If you don’t plan to use the available morphs or the face rig, you can always Clean the model (check the Clean tool in the MustardUI – Settings section of this documentation).
    Note: during clean-up, emotions and Face controls will be removed, but the standard face bones (based on weights, and colored green in viewport) will be preserved.
  • If the model crashes when rendering (Out-of-Memory errors), you can try to downgrade the textures.

Enable MustardUI Simplify and check the Check Optimization part of the UI Tutorial.

You can also enable, in the Render settings tab, Simplify. Set Max Subdivisions to 0 (only for viewport). It is always better to disable Simplify before the final rendering.

A solution can be to disable OpenGL Depth Picker in Blender settings.

Try to update both the DAZ Importer (Diffeomorphic) if you’re using it, and the MHX panel, to the latest versions:

If the MHX panel only shows an Update MHX button, please click on it to update the rig.

This is a Blender bug that prevents some panels to be loaded in some conditions. There are different ways you can try to solve this issue.

  • Delete the current MHX panel add-on (not disable), restart Blender and install the add-on from the alternative download link.
  • If the above doesn’t work, there is a workaround. Even if the panel is not available, the add-on operators should still be available to be called manually. Select the Armature and press F3 in viewport. Then search for “mhx snap”. On the left of the white name, there is the name of the operator in grey. Select the fk or ik depending on which mode you want to switch to. Repeat this for all 4 limbs if you needed.
  • As last resort, try to append everything in a clean blend file.

You can remove bone limitations on Location and Rotation in the MHX panel, using the Limits button.

If you don’t see this button, please update to the latest MHX runtime.

Remember to enable the Pose Library addon in the Blender Preferences. This addon should be available by default, but it might be not enabled.


Most of the time this issue is related to out of memory issues. In order to check this, you can try with a resource manager (for instance Ctrl+Alt+Canc in Windows), and see the values of VRAM/RAM used.

VRAM Issues

If you are having out of VRAM issues, you can try downgrading the textures to 2K. A batch conversion tutorial is available on the blog.

RAM Issues

When a Windows application fills the RAM completely, it will start using the committed memory, which is a “temporary” memory allocated on the disk. Although this is highly inefficient and slow, it prevents hard crashes of Blender due to out of RAM issues. Especially if you have heavy scenes (e.g. with a lot of particle or volumetric effects), you will see a crash as soon as Windows fills the Committed memory completely. Sometime freeing some space on your Windows main disk might help adding more Committed memory.

Be sure to check the Eevee rendering settings, enabling Screen Space Reflections (SSR) and, in the same SSR menu, enable Refraction.
For better results, also disable Half Res Trace.

Check the Eevee rendering settings. Under the Hair tab, switch Hair Shape Type from Strand to Strip.

Note that this might affect performance, so it is better to apply this setting only for renderings.

If you see some graphic artifacts on the skin like in the image, disable GPU Subdivision.

Unfortunately, this is a Blender bug and there is no known solution except disabling GPU Subdivision.


The problem is called Terminator Effect (google it). When you render, subdivider modifiers are activated and the shadows should be smoothed. You can preview this enabling the Subdivision modifier in the mesh you want to test. Note that this modifier can impact the performance of your viewport rendering. Thus, after you test its behaviour, remember to disable it (only in the viewport)!

Another solution might be to enable the Shadow Terminator Offset.

If in Cycles you ses odd looking normals (as in the image), be sure that the Eevee Optimized Normals is disabled. In fact, Eevee Optimized Normals are not to be used in Cycles and should be disabled when switching engine to Cycles.

This problem is due to the intersecting mesh with alpha textures. It affects only Cycles renderings. To solve it, increase the default value of Transparency max bounces in the rendering settings. A x3 value, from 8 to 24, should be sufficient.

Note: at the moment, a Blender bug is affecting some hair: https://developer.blender.org/T96537
The only solution at the moment is to use Blender 3.0 instead of Blender 3.1.

This usually happens when the memory needed for the rendering is bigger than the VRAM (if the GPU is used) and/or RAM allocable by Blender on your system. Check the followings:

  • sometime it happens that a lot of the memory is allocated by the viewport shading mode (Eevee for viewport shading previews). You can try if this is actually allocating too much memory going into solid mode, saving your Blender file and restarting Blender. After that, without changing viewport mode, just start the rendering;
  • consider that watching videos, playing videogames or even browsing might need memory which are not available to Blender. Try to close all software except Blender, and restart the rendering;
  • if you are using particle hair, decrease the number of Render Amount in the Children tab of the particle hair settings, the Render steps (9 might be too high, 7-5 might be nice) and the Curves additional subdivisions (0 should be fine in 99% of the applications);
  • if you have an Nvidia GPU, use Nvidia Studio Drivers instead of the Game Ready drivers;
  • try to switch to CPU rendering instead of GPU.

Note that, even if the error code is usually thrown as a CUDA error, it might be a combination of lack of GPU RAM and standard RAM. If the above does not help, you can try to downscale the textures you are using (see the tutorial) or simply consider an upgrade of your PC. If you are not sure, before buying new parts, you can ask for help on my Discord help.

This is due to the hight count of hair used by the model, combined with the heavy computational requirements of shaders settings like Principled Hair BSDF. Try to decrease the amount of rendered hair, lowering the number of Render Amount in the Children tab of the particle hair settings.

I would like to highlight some steps that you might want to check before a final render. This can be a bit time consuming, but it can greatly improve the quality of your rendering. Thus, go in rendering preview and:

  • to avoid the terminator effect (blocky shadows), check that the subdivision modifier is actually solving the problem. To do this, see the previous FAQ “The textures are blocky/there are some strange shadows in Render preview/Render” . If not, change the lights until you solve the problem.
  • check if there are some black zones in the hair. If this is the case, increase the number of transparency bounces in the rendering options until the problem is fixed.
  • check that the subsurface scattering is not too high or too low. The first case is evident if you have weird red zones on the parts of the skin with a light behind it (from the camera point of view). The second case is the inverse: you won’t see any red zone on the skin even if the light behind it is very powerful. Change the values of Subsurface Scattering and Translucency in the UI until you obtain a satisfactory result.
  • (optional) check that the denoiser is set up in the compositor (do not use the Blender default one!). If you have an RTX card you might want to enable the OptiX denoiser: but consider that, accordingly to the official Blender documentation, this denoiser is not good for animations. You can also check the custom denoiser by zixaphir.

Modify Models

There are several approaches to include physics:

  • Firstly check MustardUI, if the Physics tab is available, some parts might already have physics available.
  • The fastest, simplest, but maybe more inaccurate and inefficient way, is to use a bone addon. A great tool is Bone Dynamics Pro (there is also a free version).
  • Prepare a mesh cage, i.e. create a mesh which enclose the part which you would like to bounce, enable cloth physics on this mesh (which should be parented to an armature bone) and then bind the mesh to the body with a Mesh Deform modifier.
    I also prepared a video to explain how to add physics easily with MustardUI. At this link you can also find a pre-prepared cage to use.
  • The last method is to use soft body. This is maybe the more complicated approach, but if correctly implemented it should give the best results. I personally use the x-muscle addon to simplify the process, which is not available for free. However, some tutorials are available online on this (google it!).

If the model is Generation 3, you can follow the instructions of this tutorial available on smutba.se.