ComfyUI nodes for UltraTex — Unleashing 2K Multi-View Diffusion for 3D Texturing (SIGGRAPH Asia 2026). Give it an untextured mesh and a reference image; it generates six 2048² texture views with FLUX.2-Klein-base-9B (or FLUX.1-dev) + the UltraTex LoRA, and bakes them into a UV texture on the mesh (GLB).

The diffusion model, VAE and text encoder are loaded with the stock ComfyUI loaders, so ComfyUI's memory management, fp8/GGUF weights and offloading all apply.

Nodes

| Node | What it does | |---|---| | UltraTex Load LoRA | Applies an UltraTex LoRA. The stock Load LoRA node can't read these checkpoints (PEFT lora_A.default keys for FLUX.2, UNO processor.*_lora keys for FLUX.1). | | UltraTex Foreground VAE Decoder | Replaces the VAE decoder with UltraTex's Foreground-Aware decoder (decoder.pt). | | UltraTex Prep (mesh + reference) | Normalises the mesh, UV-unwraps it (comfy_gpu: ComfyUI's GPU unwrapper from the Unwrap Mesh UVs node, ~6 s for 500k faces; xatlas: CPU, ~4.6 min for 500k faces; or keep_existing UVs), renders the 6 canonical G-buffer TexVerse views (world normals + masks) and aligns the reference image to the front view. | | UltraTex Sampler | UltraTex sampling: background-token dropping, normal-map + reference conditioning, UNO schedule. Outputs the 6 views (in view order front, left, back, right, top, bottom), their masks and the 2×3 atlas. | | UltraTex Bake Texture | Back-projects the views into a UV texture (two-band fusion, exposure matching, mesh hole fill: see Bake), saves output/<prefix>_XXXXX_.glb (+ albedo / ORM png) and returns a TRIMESH, the texture, an 8-view preview, the GLB path and model_3d (connect it straight to Preview 3D). |

Models

| File | Folder | Source | |---|---|---| | flux-2-klein-base-9b(-fp8).safetensors | models/diffusion_models | black-forest-labs/FLUX.2-klein-base-9B — the base model, not the distilled Klein 9B | | qwen_3_8b*.safetensors | models/text_encoders | ComfyUI Klein-9B text encoder (CLIP loader type flux2) | | flux2-vae.safetensors | models/vae | FLUX.2 VAE | | flux2/lora/dit_lora.safetensors (albedo), flux2_mr/lora/dit_lora.safetensors (metallic-roughness) | models/loras (rename, e.g. ultratex_flux2.safetensors, ultratex_flux2_mr.safetensors) | ModelScope Yibo-Zhang/UltraTex | | flux2/decoder.pt, flux1/decoder.pt | models/ultratex (e.g. ultratex_flux2_decoder.pt) | same |

FLUX.1: flux1-dev in models/diffusion_models, ae.safetensors (FLUX.1 VAE), DualCLIPLoader with clip_l + t5xxl (type flux), flux1/lora/dit_lora.safetensors and flux1/decoder.pt.

FLUX.1 settings (verified against UltraTex's own FLUX.1 model code: identical DiT prediction on the same inputs, cosine 0.9995):

Reference positions are kept at their 2048 offsets at any resolution. FLUX.1 (UNO pe='d') places the normal-map and reference tokens at h/w offsets equal to the grid sizes, which the LoRA learned at 2048. At 1024 the native, halved offsets made the model lose the target-normal correspondence (face painted on the back view, incoherent side views); the sampler now always uses the 2048 offsets (no change at 2048; FLUX.2 uses resolution-independent t-planes and is unaffected).
guidance ~2. 1 gives the colours closest to the reference but softer detail, 4 (UltraTex's own FLUX.1 default) is crisper but more saturated.
UltraTex Text Encode instead of CLIP Text Encode: ComfyUI pads FLUX.1's T5 context to 256 tokens, UltraTex uses 512 (small effect, but this matches training).

Bake

fusion = two_band (default) splits every view into a low band (Gaussian blur) and a detail band:

Low band (colour, shading) is blended across views with view-angle weights, so seams stay smooth.
Detail band comes almost only from the view that sees a texel most head-on (detail_sharpness, default 32; higher = sharper, but view switches can show), so fine detail is not averaged away. This is what keeps the extra detail when you upscale the sampler's views before the bake (any size works; edge feather, blur and depth tolerance scale with the view resolution). Views are sampled bicubically.
views_2_weight (default 0.6): second-pass views compete with their view-angle quality scaled by this factor, so they paint only what the first pass sees poorly. A second pass (especially at high denoise) never matches the first pass exactly, and mixing the two softened the first pass' detail. On the dwarf test (10 views), the front detail kept rose from 0.64 to 0.75 of what a front-only bake reaches (0.84). 1 = no preference.
exposure_match: per-view colour gains estimated on the overlaps (robust, front view anchored, at most ±15%). The model often renders some views brighter (on the dwarf test, the left view is ~25% and the top ~33% brighter than the front in linear light).
Blending happens in linear light.
hole_fill = mesh: texels no view sees (under arms, inside sleeves, under the hat brim) are filled by a smooth harmonic interpolation over the mesh surface, instead of copying the nearest seen texel (nearest).

fusion = legacy is the previous bake (one weighted blend). The per-view work of two_band (background fill, low band, visibility, sampling) runs on the GPU (CPU fallback when the GPU is out of memory). Dwarf test, 10 views upscaled to 2048 px, 4K texture: two-band ~12 s, legacy ~135 s; 10 views at 4096 px used to take ~3.5 min.

Near-white pixels in a thin band just inside each view's silhouette are ignored (the model sometimes draws the outline a few pixels inside the true one and leaves the gap white, which showed up as white fringes on sleeves / boots), and view colours are extended past the silhouette before sampling. Metallic-roughness maps use the same fusion without colour gains or linear-light conversion.

Workflow (FLUX.2)

Load Diffusion Model (klein base 9B) ─ UltraTex Load LoRA ─────────┐
Load VAE (flux2-vae) ─ UltraTex Foreground VAE Decoder ────────────┤
Load CLIP (qwen 3 8B, flux2) ─ CLIP Text Encode ("")  ─ positive ──┤
Load Image (reference, RGBA) ─┬ image ─ UltraTex Prep ── prep ─────┼─ UltraTex Sampler ─ views ─ UltraTex Bake ─ model_3d ─ Preview 3D
                              └ mask ──┘   (mesh_path)        └────────────────────────────────────┘ prep

The prompt should be empty: UltraTex is image-guided. Leave negative unconnected: with an empty prompt CFG is a no-op that doubles the cost of every step (the sampler also skips it automatically when the negative is identical to the positive).
Reference image: the object on a transparent background (or connect a background-removal mask and set mask_meaning to foreground). A front view works best.

Second pass: 4 more views at 45° / 135° / 225° / 315°

UltraTex only knows its 6-view layout, so extra views come from a second pass with the camera rig turned 45° (the object is rotated; the model still sees its canonical cameras). The second pass starts from the first pass re-projected onto the rotated rig (SDEdit, denoise ~0.6) and keeps what the first pass saw head-on (keep_mask), so both passes agree; the final bake blends all 12 views. On the test meshes this raises the surface seen head-on (view angle < 45°) from 44% to 58% (dwarf) and 60% to 75% (anime girl).

UltraTex Prep ─ prep ──┬──────────────────────────────── Sampler 1 ─ views ─┬──────────────── Bake.albedo_views
                       │                                                    │    (Bake.prep = first prep)
                       ├─ UltraTex Rotate Rig (45) ─ prep ─┬─ Render Views.prep
                       │                                   │  Render Views.source_prep  ← first prep
                       │                                   │  Render Views.source_views ← Sampler 1 views
                       │                                   │       init_views, keep_mask ─┐
                       │                                   └─ Sampler 2.prep              │
                       │                                      Sampler 2 (denoise 0.6) ←───┘ ─ views ─ Bake.albedo_views_2
                       │                                                                         Bake.prep_2 ← rotated prep

Do not connect Sampler 1's views / masks straight into Sampler 2: they are seen from the other rig (the sampler refuses misaligned init_views). Sampler 2 runs denoise x steps steps (0.6 x 25 ≈ 11).

elevation on UltraTex Rotate Rig views the side cameras from above (> 0) or below (< 0):

elevation_mode = per view (default): each of the 4 side cameras is raised by elevation, upright, no roll. This is a non-canonical camera layout, but the second pass starts from the re-projected first pass, which keeps the views consistent (tested at +20°: all four side views clean).
elevation_mode = rigid: the whole 6-camera rig is tilted, so only the front camera rises; the back camera looks from below and the left/right views are rolled sideways (the model renders those poorly).

UltraTex Custom Rig (views list) replaces Rotate Rig when you want several extra rings of views: azimuths = 45,135,225,315,45,135,225,315 and elevations = 20,20,20,20,-20,-20,-20,-20 (azimuth 0 = front, 90 = right side, 180 = back, 270 = left side; a single elevation applies to all). Every group of 4 views is one second pass: each view goes to the closest front / left / back / right slot and is rendered exactly at its azimuth / elevation (no roll). The node outputs a list (one prep per group), so Render Views and Sampler 2 run once per group by themselves, and UltraTex Bake takes the whole list on prep_2 / albedo_views_2. Time grows with the number of groups (one second pass each).

Metallic-roughness

The same UltraTex Sampler predicts metallic-roughness when its model carries the MR LoRA (flux2_mr/lora/dit_lora.safetensors) instead of the albedo one. Branch the diffusion model:

Load Diffusion Model ─┬ UltraTex Load LoRA (ultratex_flux2)    ─ UltraTex Sampler ─ views ─ Bake.albedo_views
                      └ UltraTex Load LoRA (ultratex_flux2_mr) ─ UltraTex Sampler ─ views ─ Bake.metallic_roughness_views

Both samplers share the same VAE (FLUX.2 foreground decoder), positive conditioning, prep and seed. The bake writes a glTF ORM texture (G = roughness, B = metallic). MR maps are low-frequency, so the MR sampler can run at 1024 even when albedo runs at 2048.

Performance (RTX 3080 Laptop 16 GB, fp8 Klein base 9B, 25 steps, ComfyUI dynamic VRAM)

Anime-girl test asset (T-pose; bulkier objects have more foreground tokens and are slower):

| resolution | tokens in the DiT | sampling | whole workflow (prep cached, 4K bake) | |---|---|---|---| | 1024 | ~10k | 118 s (4.7 s/step) | ~2.3 min | | 2048 | ~32k | 472 s (18.9 s/step) | ~9.6 min |

Sampler optimisations (on by default):

attention = ultratex_sparse — UltraTex's block-sparse top-k attention (vendored SLA Triton kernel, the attention UltraTex is trained with). 4.4x faster than flash-attention at 64k tokens. At ~32k tokens the linear layers dominate (~37 TFLOPs bf16, compute bound), so the step gain is ~20%; for bulky meshes at 2048 (60k+ tokens) attention dominates and the gain is much larger.
token_chunk = 8192 — DiT MLPs computed in token chunks (identical result up to bf16 rounding), so activation memory stays small and the whole model stays on the GPU at 2048.
CFG is skipped when the negative equals the positive (empty prompts).
memory_factor = 0 (auto) sizes the VRAM reserve for the chunked activations.

Other levers: fewer steps (time is linear in steps), sparse_topk 0.1 (faster attention, may cost detail), run the metallic-roughness sampler at 1024.

Where the time goes (2048, bulky 500k-face character, 62k DiT tokens, ULTRATEX_PROFILE=1): single-block linear1 ~16 s, linear2 ~7 s, double blocks ~8 s, attention ~12.5 s per step, i.e. ~40-45 s per step. The matmuls run at ~75% of the GPU's burst bf16 rate; on laptops sustained runs are limited by thermal throttling (observed: 87 °C, SM clock 1110 of 2100 MHz), so cooling / power mode matters more than any remaining software setting. Set the environment variable ULTRATEX_PROFILE=1 before starting ComfyUI to log this breakdown for every step (it adds CUDA syncs, so leave it off normally).

Notes

Camera / normal conventions were reverse-engineered from the G-buffer TexVerse samples: 6 views at distance 1.67, 35.5° FoV, mesh centred and scaled to a bounding radius of 0.5, world normals stored as Blender (x, y, -z).
UltraTex code is MIT licensed. ultratex_core/sparse_attention/ is UltraTex's vendored copy of SLA (Apache-2.0, see its LICENSE.txt and PATCHES.md); it needs Triton (triton-windows on Windows). Without it the sampler falls back to ComfyUI's attention.