Filter
Exclude
Time range
-
Near
んー流石にseedvr×4はなんか無理やり感出ちゃうので×2までにしよう…
133
Replying to @LemureHead
This one, I used a combination of four open source models. Qwen 2511 edit, Klein 9B, and Z-Image, along with the SeedVR upscaler, all through the Comfy UI backend. Sometimes when i'm lazy I use grok imagine as well, but its pretty lowrez.
1
189
May 14
Replying to @ciguleva
I'd add in seedvr 2.5 for local upscaling (images and videos)
2
417
Replying to @0xCheshire
用seedVR这种图片放大,或者随便什么其他本地的图像编辑模型去小小的微调编辑一下,这东西很容易去掉的。毕竟基本都是重新采样生成的。
18
13,367
Object Swapping🔁 👗👑🧥 on Flux-2-klein-9b This simple flux-2-klein-9b flow to swap objects using a reference image. It’s pretty smooth. It uses SAM2 for the segmentation and SEEDVR to push the final result to 4K. detailed post workflow: reddit.com/r/comfyui/comment…
1
23
2,187
ComfyUI「flux-2-klein-9b」ワークフロー、参照画像によるオブジェクト入れ替え機能。SAMセグメンテーションとSEEDVRで4Kアップスケールを実現。ベース画像、参照画像、オブジェクト名の入力で利用。プロンプト自動生成。SAM2からComfyUI内蔵SAM3.1への更新も指摘。 #ComfyUI #AI生成 URLはリプ⬇️
1
1
57
4,129
Replying to @fabdreamai @peterom
TECH STACK The open source stack: → Z-Image Turbo — keyframes → LTX 2.3 Pro — video generation → SeedVR — 4K upscaling → Claude Code — AI agent orchestration → Editing: Premiere Pro CapCut → Music: "Mad Priest" by @RokNardin Hardware: RTX 3090 Mac Mini M4.
2
2
386
Replying to @RimasXYZ @replicate
We've heard people say it beats SeedVR at 1/10th of the time, so it is at least competitive.
2
31
you can use seedvr locally for free too
1
48
ComfyUIのアップスケーリングモデル、Flux2.KleinとSeedVR2を比較。 SeedVRは現代的画像の品質向上、キャラの一貫性維持に強みを持つ。 Flux2.Kleinは詳細の再生成に優れる一方、顔構造・表情変化の課題点。 この弱点は専用Consistency LoRAで補完可能。各モデルに得意分野あり。 #ComfyUI #AI画像生成 URLはリプ⬇️
1
5
541
For single image seedvr is nice, but for video I meter managed to get something useful from it besides being painfully slow.
1
2
197
Replying to @notiiivy
t2I Prompt generated using QwenVL-235B-A22B: 1) Gemini Nano Banana Pro SeedVR 7B Upscale 2) Z-Image DyPE SeedVR 7B Upscale 3) Flux.Krea SeedVR 7B Upscale 4) ChatGPT Images V2 SeedVR 7B Upscale 5) Qwen Image 2512 Q8GGUF (The worst so did not even try the upscale.)
1
1
170
31 Dec 2025
✨Wishing you all a fantastic New Year.✨ My workflow for 2K26: Midjourney Magnific Grok LightRoom Fal/SeedVR ☺️
4
1
19
881
31 Dec 2025
Replying to @underwoodxie96
Example: Image 1 = input image 2 = output I had Claude write me a small script when given an mp4 to pack into a collage and then a separate script to unpack the nano banana pro outputs into single frames that I upscale as a video sequence with SeedVR
2
8
1,029
#WomanWrestler #AIart #AIWrestler #HighWrestlingBoots #Calves  only ZIT seedvr. Portrait of a woman in a softly lit outdoor & indoor setting.
2
8
1,413
#WomanWrestler #AIart #AIWrestler #HighWrestlingBoots #Calves ⬅️nano banana pro ➡️ZIT seedvr A female wrestler sits posed on a white cylindrical platform against a plain white studio background.
1
9
87
5,254
⬅️nano banana pro ➡️ZIT seedvr Professional wrestler in wrestling ring, crouching pose, smiling, knee-high boots, arena crowd background
5
18
1,738