This is legitimately the craziest thing I have ever seen in my entire life Google deepmind is the best ai lab on planet earth right now

Another exciting Wan 2.2 model for you, for when speed matters most: replicate.com/wan-video/wan-… image-to-video and text-to-video 💥 480p: 5 seconds at 24fps in 5 seconds, $0.0125 per video 720p: 5 seconds at 24fps in 12 seconds, $0.025 per video

when you think your mvp is too simple, just know there's a skincare app making $150k monthly and it's basically 'take selfie, buy products" so i looked into how thea is actually doing this they mix ai face analysis with basic skincare advice but make it feel like getting tips

Ideogram Character + Inpainting adapts well to the given image style

Japan is shaking. 😭 AI can now create full anime scenes from a single prompt, with voice, motion, backgrounds, everything. Goodbye Anime studios and months of production. Tutorial + some crazy examples: 🧵

Coca Cola offers $750k for this ad the guy got it done for a few dollars using Higgsfield AI here are 8 raw clips with prompts:

Here me out: Genie 3 World that live processes into a Gaussian splat as you move. So exploring actively builds the 3D environment. No more 1-minute history limit, just tangible exploration that stays consistent.

We just dropped Fast Wan 2.2 14B⚡️ Wan 2.2 quality in just 6 steps < 30s for 3.5s video

> Remove makeup of this person replicate.com/black-forest-l…

🚀 New SOTA Alert! From Gen-2, 3D generation is taking on the challenge of object-level 3D scanning! Early access start soon🚘

chinese labs are already working on a budget Genie 3 called Yan which can generate interactive videos in real-time at 1080p and 60fps:

Introducing 𝗦𝗧𝗿𝗲𝗮𝗺𝟯𝗥, a new 3D geometric foundation model for efficient 3D reconstruction from streaming input. Similar to LLMs, STream3R uses casual attention during training and KVCache at inference. No need to worry about post-alignment or reconstructing from scratch.

I've built a demo for the ultra-fast high quality HyperNoise Sana Sprint 0.6B! 🔥 thanks for the team Google for open sourcing the technique & model! kudos to Luca Eyring, Nataniel Ruiz and team for making that happen 🤩

Great summary of the latest image & video diffusion models! Our #SIGGRAPH2025 course spans real-world uses to techniques like acceleration & flow matching. Slides: geometry.cs.ucl.ac.uk/courses/diffus… w/ Niloy Mitra, Or Patashnik, Daniel Cohen-Or , Paul Guerrero, Minhyuk Sung

IT'S OUT! 🚀 MoDA: Multi-modal Diffusion Architecture for Talking Head Generation finally a talking head: open source 🏋️ fast ⚡ portrait + audio-driven 🧑‍🎨🎧 with emotion control (and yes, i built an inference system + Gradio, generate in < 15s on Hugging Face spaces 🤗)

Qwen Image Edit. 3 seconds, $0.03. replicate.com/qwen/qwen-imag… Pruna AI do the impossible. > Make the text 3D and floating on a city street

Qwen has solved image editing $0.03, 3 seconds per edit on Replicate Everyone go find a different problem to solve lol replicate.com/p/182jkcqmr9rm…

🧠Dreaming Body and View in First Person🧠 #EgoTwin is a 😎egocentric world model😎 that jointly predicts & generates **egocentric video** and **human motion** in a view consistent and causally coherent manner. - Project: egotwin.pages.dev - Paper: arxiv.org/abs/2508.13013

try it replicate.com/lucataco/multi…

Add a Girlfriend with Qwen-image-edit