🔴 Новая модель генерации видео Grok Video 1.5. Сравниваю ⚡️Grok Video 1.5 c 🔥Seedance 2.0

😎 Subscribe to the Telegram channel about neural networks 👉 https://t.me/neyr0graph 🧲 Subscribe to a chat with like-minded people 👉 https://t.me/neyrodev ❤️ Subscribe to Boosty 👉 https://boosty.to/neyrograph 🔥👑 WAEVY AI - the best platform for generating images and videos - https://www.weavy.ai/?ref=sergey14 🔥 The main and most famous one is VBench from the Vchitect team https://github.com/Vchitect/VBench 00:00 - Introduction: Overview of the new Grok model Video 1.5 00:40 - Blogger ratings and comparison with Seedance 2.0 01:06 - Visual benchmark and evaluation tools (V-Bench) 01:33 - Technical capabilities and Grok and Seedance 2.0 features 02:22 - Model parameters: duration, resolution, and engine 02:40 - How to generate cheaper in WeavyAI: a useful hack 03:03 - A detailed analysis of working with references, file types, and links 04:05 - Comparison of generation costs and the benefits of upscaling 05:33 - Categories for testing neural networks 05:57 - Real model ratings in independent arenas 06:42 - Overview of the workflow in the Weavy AI interface 07:33 - Benefits of the Boosty community 08:20 - Test 1: A dynamic fight scene between two characters 09:50 - Test 2: Lip sync in UGC format in Russian 11:37 - Test 3: Dialogue between elderly people (checking choreography and emotions) 12:47 - Test 4: Using a character mesh for Cinematic atmosphere 14:43 - Test 5: Passing facial censorship using a complex prompt with a skier 17:37 - Test 6: Censoring exposed body parts and the behavior of physics in the frame 19:56 - Test 7: Marketplaces and product offerings (working with text on a rotating jar) 23:27 - Useful resources: Telegram channel on neuroart and the "NeuroDely" chat 24:29 - Announcement of the next video about AI directing and the 180-degree rule 25:04 - Final conclusions: Pros and cons of Grok 1.5 and Seedance 2.0 In the new video, I use my own examples to demonstrate the capabilities of the new Grok 1.5 video model compared to the current leader, Seedance 2.0. I demonstrated and explained how both neural networks handle complex tasks: generating dynamic combat, maintaining small text on marketplace products, bypassing strict facial censorship using special meshes, and creating realistic Russian lip-synch. Furthermore, I provided a detailed comparison of generation costs and practical tools for optimizing AI content production costs. I explained how to generate content more cheaply in Weavy AI: a useful hack available on Boosty.