━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━🔼 SeedVR2 Batch Upscaler — Sleep On It, Wake Up 4KDrop a folder, come back to 4K━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━ComfyUI batch image upscaler powered by SeedVR2, the SOTA diffusion-based restoration model from ByteDance. Load an entire folder of QC-approved images, set the batch number, click Run once, and watch them upscale to 4K in a single session. No per-image re-queuing, no repetition—sequential images feed through the same upscaler instance to your output folder, all with automatic date-stamped organization. Built for production stock photography, concept art detail recovery, and QC'd generation batches.━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━✨ Features✅ Batch Folder Iterator — Place 10–100 images in input folder, set the batch number to match, all process sequentially without re-loading model✅ Whole-Folder Workflow — ImageIterator node handles file discovery, sorting, and index auto-increment✅ SeedVR2 7B fp16 (default) — Diffusion-based upscaler, SOTA quality; fp8 and 3B variants available for speed/VRAM trade-off✅ 4K Short-Edge Target — short edge upscaled to 4096, long edge capped at 4096 (1024×1024 → 4096×4096; 1280×1600 → 3276×4096). The effective upscale factor depends on the input's aspect ratio.✅ Block-Swap Ready — Pre-configured with blocks_to_swap=36 + CPU offload + SDPA attention for 16GB VRAM fit (requires ~33GB system RAM)✅ Automated Output Organization — Results saved to output/ folder with automatic date stamp [YYYY-MM-DD]✅ Torch.compile Optional — Node included but disabled by default; enable if you have triton-windows for inference speedup✅ Metadata-Safe Export — PNG output; embedded workflow metadata has no machine paths or language-specific notes━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━📦 Required Models (3 files, ~7-16 GB depending on variant)Primary (default — recommended)• seedvr2_ema_7b_fp16.safetensors — Main diffusion upscaler (highest quality, fits 16GB with block-swap)• ema_vae_fp16.safetensors — VAE codec for image reconstructionAlternative Variants (choose one — do NOT mix in same batch)• seedvr2_ema_7b_fp8_e4m3fn.safetensors + ema_vae_fp16.safetensors — Faster, near-identical quality, lower VRAM• seedvr2_ema_3b_fp16.safetensors + ema_vae_fp16.safetensors — Fastest option, fits 16GB without block-swap━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━⬇️ Download Links (verified HuggingFace repositories)7B fp16 (default — recommended for quality)• seedvr2_ema_7b_fp16.safetensors — <a target="_blank" rel="ugc" href="https://huggingface.co/numz/SeedVR2_comfyUI/blob/main/seedvr2_ema_7b_fp16.safetensors">https://huggingface.co/numz/SeedVR2_comfyUI/blob/main/seedvr2_ema_7b_fp16.safetensors</a>• ema_vae_fp16.safetensors — <a target="_blank" rel="ugc" href="https://huggingface.co/numz/SeedVR2_comfyUI/blob/main/ema_vae_fp16.safetensors">https://huggingface.co/numz/SeedVR2_comfyUI/blob/main/ema_vae_fp16.safetensors</a>7B fp8 (faster, lower VRAM)• seedvr2_ema_7b_fp8_e4m3fn.safetensors — <a target="_blank" rel="ugc" href="https://huggingface.co/numz/SeedVR2_comfyUI/blob/main/seedvr2_ema_7b_fp8_e4m3fn.safetensors">https://huggingface.co/numz/SeedVR2_comfyUI/blob/main/seedvr2_ema_7b_fp8_e4m3fn.safetensors</a>• ema_vae_fp16.safetensors — <a target="_blank" rel="ugc" href="https://huggingface.co/numz/SeedVR2_comfyUI/blob/main/ema_vae_fp16.safetensors">https://huggingface.co/numz/SeedVR2_comfyUI/blob/main/ema_vae_fp16.safetensors</a>3B (fastest, native 16GB fit)• seedvr2_ema_3b_fp16.safetensors — <a target="_blank" rel="ugc" href="https://huggingface.co/numz/SeedVR2_comfyUI/blob/main/seedvr2_ema_3b_fp16.safetensors">https://huggingface.co/numz/SeedVR2_comfyUI/blob/main/seedvr2_ema_3b_fp16.safetensors</a>• ema_vae_fp16.safetensors — <a target="_blank" rel="ugc" href="https://huggingface.co/numz/SeedVR2_comfyUI/blob/main/ema_vae_fp16.safetensors">https://huggingface.co/numz/SeedVR2_comfyUI/blob/main/ema_vae_fp16.safetensors</a>Installation Path: • Diffusion models: ComfyUI/models/diffusion_models/• VAE files: ComfyUI/models/vae/━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━🧩 Required Custom Nodes⚠️ Manual Install Step (do this first — it's the only one not in ComfyUI Manager):Image_Anything (Batch Folder Iterator) — NOT in ComfyUI Manager registry1. Open terminal/PowerShell in your ComfyUI root directory2. Navigate to custom_nodes: <code>cd custom_nodes</code>3. Clone the repo: <code>git clone https://github.com/ComfyUI-Kelin/ComfyUI_Image_Anything.git</code>4. Restart ComfyUIThen install these via ComfyUI Manager (search → install):• ComfyUI-SeedVR2_VideoUpscaler (numz) — canonical SeedVR2 node pack, in Manager registry, searchable by name - Install via Manager: Manager → Install Custom Nodes → search "SeedVR2" → select numz's ComfyUI-SeedVR2_VideoUpscaler → Install• was-node-suite-comfyui (ltdrdata) — Image Save node (PNG/WebP export); MIT license, in Manager registry - Install via Manager: search "was-node-suite" → installVerify Installation: After restart, load this workflow in ComfyUI. If nodes resolve (no red outlines), you're good to go.━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━🚀 How to Use1. Prepare Images — Place your QC-approved images (PNG/JPG, any resolution 256–4096px) in ComfyUI's <code>./input</code> folder2. Load This Workflow — Open SeedVR2_Batch_Upscale_v1.1.json in ComfyUI3. Run It — Set the batch number next to Run to the number of images you want, then click Run once (older ComfyUI builds label this button Queue) - Each execution advances the ImageIterator by one file, so a batch of 10 walks 10 images. Clicking Run repeatedly does not stack on current frontends — it ignores presses until the current run finishes, so use the batch number.4. Monitor Progress — Watch the ImageIterator node report "current_index / total_count" as it walks the folder5. Collect Output — Check ComfyUI's <code>./output</code> folder for upscaled images, organized by date [YYYY-MM-DD]Example Workflow:- Input folder: <code>ComfyUI/input/</code> (contains 5 PNG files)- Set the batch number to 5 and click Run once- Wait ~1 minute per image on the default 7B fp16 config (measured — see the benchmark table below); slower if your system RAM is tight- Output appears in <code>ComfyUI/output/[YYYY-MM-DD]/</code>, named from the Image Save prefix (default <code>Up VR2 [time(...)]</code> plus a counter — edit the prefix widget on the Image Save node to change it)━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━⚙️ Settings &amp; ParametersModel Loading• Model Path: Automatically detects from ComfyUI/models/diffusion_models/ (VAE loads from ComfyUI/models/vae/)• Device: CUDA (GPU); falls back to CPU if CUDA unavailable• Offload Device: CPU (system RAM) — keeps VRAM leanUpscale Output• Resolution: 4096 — target for the SHORT edge; aspect ratio is preserved• Max Resolution: 4096 — cap on the LONG edge; if an image exceeds it after the short-edge target is applied, both dimensions scale down proportionally. This keeps VRAM predictable on tall or wide images.• Batch Size: 1 per queue (sequential processing)• Color Correction: "lab" (perceptually-aware color preservation)• Attention Mode: "sdpa" (scaled-dot-product attention — memory-efficient)Block-Swap (VRAM Management)• blocks_to_swap: 36 (default, pre-configured for 16GB VRAM) - Swaps 36 UNet blocks to system RAM to fit model on GPU - Requires ~33GB system RAM; 64GB recommended for smooth performance - If you have less RAM, switch to 7B fp8 or 3B variantTorch.compile (Optional Speedup)• Status: DISABLED by default (node present, not connected)• Reason: Requires triton-windows, which most Windows users lack → would error on first run• How to Enable (if you have triton): Reconnect the <code>SeedVR2TorchCompileSettings</code> node output → the <code>torch_compile_args</code> input on <code>SeedVR2LoadDiTModel</code>. The same input also exists on <code>SeedVR2LoadVAEModel</code> if you want to compile the VAE too. Can add ~20–30% speed boost, but only if dependencies are met.Seed &amp; Noise• Seed: 42, with control_after_generate set to "fixed" — every run is reproducible. Set the control to "randomize" if you want variation between runs.• Input Noise Scale: 0.05 · Latent Noise Scale: 0.0 — low values preserve detail and prevent hallucinationVAE Tiling (if image &gt;2048px)• Encode Tiled: Enabled (1024px tiles, 128px overlap)• Decode Tiled: Enabled (same tiling)→ Prevents OOM on large inputs━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━💡 Performance TipsVRAM &amp; System RAM Honesty• 7B fp16 + swap36 (default, shipped) — ~16GB VRAM, ~33GB offloaded to CPU. Highest quality. Requires 64GB system RAM for smooth batch processing; 32GB will thrash and be very slow. Measured: ~62s/image (steady-state, 17-image production batch, 1280×1600→3276×4096)*• 7B fp8_e4m3fn — ~9GB VRAM, ~18GB offloaded to CPU. Faster inference (~30% speedup estimated, not yet benchmarked). Quality near-identical to fp16. Fits 32GB system RAM comfortably.• 3B fp16 — ~16GB VRAM, no offload needed. Fastest option (not yet benchmarked). No block-swap overhead. Fits 16GB natively.*Measured on RTX 5080, 7B fp16 variant only, back-to-back queue (model already loaded). fp8/3B timings not yet benchmarked — estimates only.Practical Guidance• For highest quality + 64GB system RAM: Use 7B fp16 (default config)• For speed + 32GB RAM: Switch model to 7B fp8 (same setup, different checkpoint file)• For minimal VRAM/RAM: Use 3B variant (nearly as good, no block-swap delays)• Batch 10+ images together to amortize the one-time model load• Avoid running alongside heavy generation (e.g., Krea2 gen) — too much total VRAM pressure• Input image quality: Sharp, well-lit images upscale better than low-contrast originals (expected for all upscalers)━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━🔗 My Workflow SuiteI maintain a growing library of ComfyUI workflows. Check them out:• LTX-2.3 Image-to-Video — Lock-camera i2v with auto motion prompt (QwenVL)• Krea2 Turbo Dual-Mode — Fast diffusion gen (text or image-to-image)• Z-Image-Turbo — Another fast gen option with auto-prompt: <a target="_blank" rel="ugc" href="https://civarchive.com/models/2742740/z-image-turbo-qwenvl-dual-mode-auto-prompt">https://civarchive.com/models/2742740/z-image-turbo-qwenvl-dual-mode-auto-prompt</a>• SeedVR2 Batch Upscaler — This workflow (batch folder upscale to 4K)Find LTX-2.3 and Krea2 Turbo on my Civitai profile page. More coming soon — follow for updates!GitHub Mirror: <a target="_blank" rel="ugc" href="https://github.com/Thinni63/comfyui-workflows/tree/main/seedvr2-batch-upscale">https://github.com/Thinni63/comfyui-workflows/tree/main/seedvr2-batch-upscale</a>━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━🔄 Changelogv1.1 (2026-08-17)• Fixed: SeedVR2 was never actually connected. The Preview node carried a phantom output slot that ComfyUI drops when loading the file, which cut the link into SeedVR2, while a stale link ID made the Save node read the preview image instead. Net effect: saved files came out identical to the input. Thanks to daze45230 for the detailed report.• Changed: SeedVR2 is no longer hidden inside a subgraph. The model loaders, upscaler, and torch.compile node now sit directly on the canvas, so every setting is visible and editable without opening a container — and a subgraph-capable ComfyUI build is no longer required.• Improved: Node layout reorganized into labelled groups (start guide / run settings / model setup / pipeline).• Settings and model requirements are unchanged — drop-in replacement for v1.v1 (2026-07-03)• Initial release.━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━📝 Notes &amp; AI Disclosure• AI-Generated Example Images — All provided examples are AI-generated via Krea2 + upscaled with this workflow• Hardware Tested — Verified on RTX 5080 16GB VRAM, Windows 11, CUDA 12.1+• ComfyUI Version — No subgraph support required as of v1.1 (the graph is flat). Any reasonably recent ComfyUI with the three custom node packs installed will load it.• Model Weights License — SeedVR2 weights are NOT distributed with the workflow; you download them separately from HuggingFace (see Download Links above)• Metadata Safety — This workflow JSON has no machine paths or language-specific notes; it's safe to share and distribute━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━⭐ Found this useful?• Like if it saved you time• Comment your results — I read every one• Follow for new ComfyUI workflows, all tested on 16 GB VRAM━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━Roadmap TeaserA video-batch variant of this upscaler is coming soon. Stay tuned!━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━⚖️ Model Attribution &amp; LicensingSeedVR2 (ByteDance Seed Team)• License: Apache License 2.0 — <a target="_blank" rel="ugc" href="https://huggingface.co/ByteDance-Seed/SeedVR2-7B">https://huggingface.co/ByteDance-Seed/SeedVR2-7B</a>• Attribution Required: "SeedVR2 by ByteDance Seed Team, licensed under Apache License 2.0."• Safetensors Conversion: Hosted at <a target="_blank" rel="ugc" href="https://huggingface.co/numz/SeedVR2_comfyUI">https://huggingface.co/numz/SeedVR2_comfyUI</a> for ComfyUI-compatible format• Free for commercial and non-commercial use (see license for full terms)Custom Nodes• ComfyUI-SeedVR2_VideoUpscaler (numz) — Check repository for license• was-node-suite-comfyui (ltdrdata) — MIT License• ComfyUI_Image_Anything (ComfyUI-Kelin) — MIT LicenseThis Workflow (JSON Configuration)• Original work by TP_AI_63 (Civitai) / Thinni63 (GitHub)• Shared under MIT License; credit appreciated• Model weights must be downloaded separately (not included)All example outputs are AI-generated via Krea2 generation and SeedVR2 upscaling.

SeedVR2 Batch Upscaler — Sleep On It, Wake Up 4K

🆕 What's new in v2✅ Any folder, sub-folders included — type your image folder in one box. No more copying files into ComfyUI/input.✅ Results keep your file names and sub-folders — in the folder you choose.✅ Stop-proof — power cut or Ctrl+C? Press Run again: finished images are skipped, nothing is redone, and the queue stops itself when everything is done.✅ 2K or 4K file (longest edge) and three VRAM files: 16 GB / 12 GB / 8 GB — see the honest numbers below.✅ No WAS node needed anymore — only ComfyUI_Image_Anything + SeedVR2_VideoUpscaler.✅ All 6 workflows (16/12/8 GB x 2K/4K) are inside the ZIP "SeedVR2_Batch_Upscale_v2_ALL_6_workflows_and_tools" — extract it first. The single JSON shown above is the 16 GB 2K file.✅ Helper scripts (plain Python): skip images that are already big, copy your prompts into the results, convert to JPEG/WebP (4K PNG 12–18 MB → JPEG 2–3 MB), before/after sheets, batch report.✅ Fixed seed (42): same input, same output.🔼 SeedVR2 Batch Upscaler — drop a folder, come back to 4KWhole-folder image upscaler for ComfyUI powered by SeedVR2 (ByteDance). Set the batch number to the number of images, press Run (older ComfyUI: Queue) once, go to sleep.🚀 Quick start (5 steps)<ol><li>Install nodes: ComfyUI_Image_Anything (git clone <a target="_blank" rel="ugc" href="https://github.com/ComfyUI-Kelin/ComfyUI_Image_Anything.git">https://github.com/ComfyUI-Kelin/ComfyUI_Image_Anything.git</a> into custom_nodes, then restart ComfyUI) + ComfyUI-SeedVR2_VideoUpscaler. Download models: seedvr2_ema_7b_fp16 (16 GB file) or seedvr2_ema_3b_fp16 (12/8 GB files) + ema_vae_fp16.</li><li>2. Open the file for your card and size: SeedVR2_Batch_Upscale_v2_&lt;16GB|12GB|8GB&gt;_&lt;2K|4K&gt;.json</li><li>3. Fill two paths: Image Iterator → folder_path, Processed Image Check → output_root.</li><li>4. Set the batch number next to Run to the number of images (or more), press Run (Queue in older versions) once.</li><li>5. Come back later. After a stop, press Run again with the same paths.</li></ol>📊 Measured on RTX 5080 Laptop 16 GB / 64 GB RAM, 4K files, 12 images each• 16GB file (7B fp16): 66 s/image · VRAM peak 12.6 GB · +29 GB system RAM• 12GB file (3B fp16): 73 s/image · VRAM peak 15.6 GB · +9 GB RAM• 8GB file (3B fp16, block swap): 46 s/image · VRAM peak 14.3 GB · +11 GB RAM• v1.1 setting for comparison: 45 s/image (single pass, 3 images)⚠️ Please read• The 12 GB and 8 GB files were run on a 16 GB card — not tested on real 12 GB / 8 GB cards. With ~7 GB of VRAM deliberately held back, the 8GB_4K file finished 6/6 images without out-of-memory (~44 s/image). 12GB_4K is experimental: not proven with ~11 GB free (the card peaked at ~16 GB). Nothing in the files enforces a VRAM limit — if you run out of memory, use a smaller file.• The 16 GB (7B) file needs a lot of system RAM (+29 GB). With under ~40 GB free it can freeze or restart a PC. 32 GB RAM → use the 3B files.• 2K files are one pass: about 13 s/image (3B, 12/8 GB files) and 24 s/image (7B, 16 GB file) on the 16 GB test card; the 3B 2K files also passed a stop/resume test. 4K is the maximum — 6K/8K crashed the test PC.• Best on photo-like images. On painterly scenery (a tree, a sea wave in our tests) 100% crops showed outlined leaves and spiky edges — check at 100%.• Compared on the same images, 100% crops of v2 were slightly sharper than the v1.1 setting (hair, lashes, fabric). Text inside images stays wrong in both. v2 (two passes) is slower than v1.1.• Outputs are PNG; the workflow's saver does not write the original prompt — use copy_metadata.py before you post.Free to use. SeedVR2 and its models keep their own licences (not for resale on their own).

Files with SHA256: 2218483f7cc00a0396d4ae93424cd8cf870cc0b4d566e4c02521c9ea00a0434b