Y
HN Search
Hacker News Search
new
|
comments
|
top
|
jobs
laiwuchiyuan
searching PlanetScale…
1.
▲
2.
▲
3.
▲
4.
▲
5.
▲
6.
▲
4 ms
·
1.
▲
Show HN: Wan-Animate – Motion Transfer for Any Character
(wananimate.net)
1 points
by
laiwuchiyuan
1y ago
|
0 comments
2.
▲
by
laiwuchiyuan
1y ago
The results generated by Kling AI Avatar are really good and of high quality, with the only advantage being that it supports 1080P and up to a minute of video length. I testing: https://www.klingaiavatar.net/
3.
▲
Qwen-Image: AI Image Generation with Native Text Rendering
(qwenimagen.com)
2 points
by
laiwuchiyuan
1y ago
|
0 comments
4.
▲
by
laiwuchiyuan
1y ago
“Qwen‑Image: Open‑source 20 B MMDiT model with stunning text rendering and image editing. Effortlessly create bilingual posters, infographics, slides, infill edits, comics.” Go experience AI: https://www.qwenimagen.com/ Why
5.
▲
Wan AI – Wan 2.2: Leading AI Video Generation Model
(wan-ai.co)
1 points
by
laiwuchiyuan
1y ago
|
0 comments
6.
▲
Alibaba open-sources Wan2.2: 27B-param MoE video models, 720 p 24 FPS on a 4090
(wan-ai.co)
2 points
by
laiwuchiyuan
1y ago
|
1 comments
7.
▲
by
laiwuchiyuan
1y ago
Alibaba’s Tongyi Wanxiang team just dropped Wan2.2—three new video-generation checkpoints under Apache-2.0: • Wan2.2-T2V-A14B – 27 B total / 14 B active, first open-source text-to-video model built on a mixture-of-experts (MoE) diffusi