Hi there UE5 and Blender/Maya etc experts, we need a full-quality Epic Games' Metahuman (hyperrealistic, from their catalogue) converted to a single, clean glTF file that we can drop straight into a web-based A2F (audio-to-face) pipeline. Please start from the original UE5 asset, then handle the export in Blender/Maya/your choice, there you need to make sure both hyper-realistic look and working Blendshapes are kept!, so the file arrives with every nuance (as per tech possibility) of the original UE5 character intact. Facial precision is critical: the core set of blendshapes / morph targets / shapekeys must survive the trip. When we drive lipsync on the site we want the same subtle micro-expressions we see inside Unreal. Visual fidelity matters just as much. Skin, eyes, teeth, tongue, clothing, and groomed hair should look as much technically indistinguishable as possible from the source original UE5 Metahuman once re-imported. Whichever hair solution best preserves realism in .glTF export —haircards, polygons, or another proven approach—is fine as long as the result stays lightweight enough for the web. I will be fully fine to use one of the current/existing Metahumans on the hyperrealistic catalogue as starting point and reference. We will choose together who. Deliverable: • One .glTF 2.0 of max 15 - 20 MB with full geometry, materials/look&feel, textures, and the core functioning morph targets/blendshape set, ready for immediate use in browser-side animation code. We will validate by re-importing Blender and to our WebGL viewer; everything must align with the Metahuman’s original proportions, UVs, and facial performance range. If that sounds straightforward, let’s get the file moving. Do not bid if you do not feel confortable please, only experienced UE5 Riggers. Critical to keep and manage morph tagets, recreate hair and look&feel from hyperrealistic Metahuman original, and to optimize so that the .glTF is having the look, the functioning (morph targets) and still relatively light for web application (max 20 MB or so). Many thanks!