Aximo Loot
17.3K subscribers
88 photos
72 videos
15 links
https://t.me/+RA1tsb6LPFVkMWNh

Share this link with your friends for joining this channel
Download Telegram
Generate 3D models locally on your PC โ€” Hunyuan3D portable build โœ”๏ธ

A new portable version of Hunyuan3D is out, allowing you to create 3D models locally on Windows 10/11 with a one-click setup.

๐ŸŸขWhat you get:
โ€” Simple browser-based interface, easy for anyone to use
โ€” Quick setup: download the archive, run the installer script, and youโ€™re ready
โ€” Generates full 3D models from images and text prompts, with proper meshes and textures
๏ปฟ
๐ŸŸขSystem requirements:
โ€” From 3 GB VRAM
โ€” Works even on older GPUs like GTX 980
โ€” 20โ€“30 GB of free disk space


@AximoLoot
Please open Telegram to view this post
VIEW IN TELEGRAM
Please open Telegram to view this post
VIEW IN TELEGRAM
This media is not supported in your browser
VIEW IN TELEGRAM
Scientists showed what the very first rave on Earth looked like ๐Ÿ˜‚

@AximoLoot
๐Ÿ˜1๐Ÿ˜ฑ1
This media is not supported in your browser
VIEW IN TELEGRAM
Create Apple-style presentations โ€” free slide builder

A new tool turns ะพะฑั‹ั‡ะฝั‹ะต ัะปะฐะนะดั‹ into sleek, animated scenes in the style of Apple presentations.


๐Ÿ“ผ Just move objects in space, and the app automatically creates smooth, cinematic camera motion
๐Ÿ“ผ Supports 3D models, images, and text blocks
๐Ÿ“ผ Works fully offline โ€” your data stays on your device
๐Ÿ“ผ No subscriptions or hidden limits โ€” everything runs locally

@AximoLoot
Please open Telegram to view this post
VIEW IN TELEGRAM
This media is not supported in your browser
VIEW IN TELEGRAM
A game-changer for sound engineers โ€” meet SAM Audio

A new tool, SAM Audio, is being called the first neural network capable of editing sound in real time.

What it can do:

๐Ÿ”น Prompt-based recording โ€” just describe the sound you need, and the AI instantly isolates it. For example, capture only a voice on a noisy street or a single vocalist at a concert.

๐Ÿ”น Visual editing โ€” tap on a sound source in the frame, and SAM immediately extracts the corresponding audio. You can even check quality during recording.

๐Ÿ”น Time-based selection โ€” choose a segment, and the model finds and separates the desired sound into its own track.

@AximoLoot
Please open Telegram to view this post
VIEW IN TELEGRAM
This media is not supported in your browser
VIEW IN TELEGRAM
A revolution in image upscaling โ€” the Thera neural network has been released, capable of rescuing even the worst-quality photos ๐Ÿ”ป

The developers highlight its main advantage โ€” ZERO distortion in the image. You get the same photo on output, just without pixels.

๐Ÿ’ป It can be used online or installed locally as well.

@AximoLoot
Please open Telegram to view this post
VIEW IN TELEGRAM
This media is not supported in your browser
VIEW IN TELEGRAM
AI Right on Your iPhone Home Screen

โ–ซ๏ธSkye is a new kind of intelligent assistant built directly into your home screen. It continuously monitors your calendar, tasks, activity, emails, and more โ€” surfacing only what matters most, exactly when you need it.

โ–ซ๏ธFrom quick email replies to alerts about suspicious charges or instant insights about your surroundings (like nearby spots or local facts), everything important stays front and center without extra effort.

Bring AI to your home screen?

โค๏ธ โ€” Yes, this is actually useful
๐Ÿ‘ โ€” No, it just clutters the interface


@AximoLoot
โค1
Cute trend in GPT Image 2

๐Ÿ“ธ Version: GPT Image 2

We create a pencil drawing based on your photos. Upload a photo and use the prompt:

Transform this image into a pencil-style illustration. Simplify all details so it looks like it was drawn by a 10-year-old child. Use soft, random, childlike lipstick colors instead. Make everything look slightly uneven, messy, and playful, like a childโ€™s drawing. Render it as if it were drawn on plain white paper with visible lipstick strokes and rough outlines. Add cute doodles around the scene, such as stars, hearts, flowers, candies, clouds, and sparkles. Make the whole image feel cozy, warm, and cute. Keep the main subjects recognizable (a girl, a cat, a home interior, kitchen items, sunset landscapes), but simplified into basic shapes and childlike proportions.


Generated on: TrueVideo.ai.

@AximoLoot
Please open Telegram to view this post
VIEW IN TELEGRAM
Please open Telegram to view this post
VIEW IN TELEGRAM
This media is not supported in your browser
VIEW IN TELEGRAM
Figma has now become an all-in-one creation generator ๐Ÿ–Œ

The new Figma Weave app introduces the ability to build workspaces for generating images, videos, 3D models, and much more.

It works similarly to ComfyUI, but fully online โ€” which makes it really convenient.

@AximoLoot
Please open Telegram to view this post
VIEW IN TELEGRAM
This media is not supported in your browser
VIEW IN TELEGRAM
The neural network was asked to depict dog owners who are allowed to walk near a skate park, and hereโ€™s what it came up with ๐Ÿ˜

@AximoLoot
Developers, get ready: Nvidia has dropped an open-source world generation model โ€” Lyra 2.0.

๐Ÿ”˜The system turns a single image into a fully interactive 3D game world, where you can walk around, explore everything, enter buildings, and freely rotate the camera in any direction.

And yes โ€” the world no longer breaks apart or forgets what happened a second ago.

๐Ÿ”˜In other words, you can now fully walk through the Shire or step inside a Salvador Dalรญ painting without feeling like youโ€™re dreaming with a 40ยฐC fever.

@AximoLoot
Please open Telegram to view this post
VIEW IN TELEGRAM
OpenAI has released GPT-Rosalind โ€” an AI model designed for science, biology, chemistry, and drug discovery.

โบIt helps researchers analyze data faster, uncover hidden connections, and generate hypotheses to accelerate scientific breakthroughs and the development of new medicines.

โบFor now, the model is only available in preview to selected biotech and enterprise companies.

@AximoLoot
Please open Telegram to view this post
VIEW IN TELEGRAM
A guy simply took a photo of his sandwich from a pizza place, uploaded it to the new ChatGPT, and asked it to create an ad โ€” the result is in the picture ๐Ÿ‘€

The prompt was only about 6 lines long:

gradient background, soft โ€œApple-styleโ€ lighting, centered typography, a slogan, and โ€œultra-clean minimalism, 8K.โ€ Thatโ€™s it.


AI can now turn ordinary photos into full-fledged advertising posters in minutes ๐Ÿ‘


Generated on: TrueVideo.ai.

@AximoLoot
Please open Telegram to view this post
VIEW IN TELEGRAM
Please open Telegram to view this post
VIEW IN TELEGRAM
This media is not supported in your browser
VIEW IN TELEGRAM
OpenAI Releases GPT-5.5 ๐Ÿ”˜

The new model is already outperforming its biggest competitors โ€” Claude Opus 4.7 and Googleโ€™s Gemini 3.1 Pro โ€” and even beats GPT-5.4 Pro in several benchmarks.

GPT-5.5 is heavily focused on agentic workflows and usability: it better understands user intent and can handle most tasks with minimal guidance.

โ–ซ๏ธ Paid ChatGPT users are already receiving access, though the API pricing is now twice as high as GPT-5.4 โ€” $5/$30 per 1M input/output tokens.

@AximoLoot
โค1
๐Ÿ”ฅ Magic Resume builds the perfect CV in just a few minutes

A new smart editor helps you create a professional resume fast โ€” no Word or complicated templates needed.


โ€” Real-time preview of every change
โ€” Export to PDF in just a couple of clicks
โ€” Auto-save and local storage โ€” your data stays on your device
โ€” Completely free and works ะฟั€ัะผะพ in your browser on any device

@AximoLoot