Telegram каналы "SmartSoftHub"

SmartSoftHub

Public

Назар аударыңыз : Кез келген ақпаратқа, әсіресе ақша сұрауға байланысты болса, сенуден бұрын өзіңіз зерттеу жүргізу маңызды.

Қосыла алмайсыз ба? @smartapphub

5.9k Мүшелер

Жаңартылған: March 11, 2026 at 3:35 AM

SmartSoftHub

💎News From The World of Free Software & AI 🌀Welcome! Here you’ll find valuable information on the latest trending projects. #ai #software #devops #web 🎯Ads: https://adsly.me/@smartapphub 🌐Site: https://smartsofthub.us 😎Admin: @SmartSoftHubUSA

Join @smartapphub for exclusive from and software content and discussions in 1

Рейтинг

Жалпы рейтинг

#3998Өзгеріс жоқ

Тіл бойынша рейтинг

#876Өзгеріс жоқ

Санат бойынша рейтинг

#30Өзгеріс жоқ

Жазылушылардың өсуі (Соңғы 30 күн)

Барлығы: 5.9K

24 сағаттағы өсу: -25 0%

Соңғы посттар

SmartSoftHub

2026 ж. 11 нау., 01:39

📷 Photo

🏞️ IC-Custom

🛠️IC-Custom is an open-source tool for various image customization based on in-context learning.

🔰It supports the accurate placement of an object from a reference image using a mask or segmentation, and the generation of a new scene with the subject from the reference based on a text description.

🔰IC-Custom can be applied to tasks such as virtual fitting, product placement, furniture arrangement, and character creation.

🔗Links:
https://github.com/TencentARC/IC-Custom/

🏷️Tags:
#ComfyUI #imageeditor #ai

3,380

SmartSoftHub

2026 ж. 11 нау., 01:39

🎥 Video

📲Open-AutoGLM

🛠️Open-AutoGLM — a ready-made framework to turn an Android phone into a "smart agent" that understands the screen and performs actions based on a user's text request.

🔰Open-AutoGLM is built around the AutoGLM-Phone-9B model and uses it as a brain that sees a screenshot of the screen, understands the interface, and plans step by step what to press and where to enter text.

🔗Links:
https://github.com/zai-org/Open-AutoGLM
🏷️Tags :
#llm #phone #ai

3,730

SmartSoftHub

2026 ж. 11 нау., 01:39

🎥 Video

🤩 Black Forest Labs has released FLUX.2 — a new generation open-source image generator.

🛠️ This is the most powerful open model for image generation to date, combining text-to-image and editing in a single checkpoint.

🔰 The AI is already available on Hugging Face and in Flux Playground.

🔰 The architecture is built on a combination of the vision-language model Mistral-3 with a diffusion transformer — this approach gives the model contextual understanding and precise scene composition handling.

🔰 FLUX.2 can work with multiple references at once (up to 10 images), generate images up to 2048×2048, and render text with high quality.

🔰 For local deployment, NVIDIA and ComfyUI have jointly prepared an optimized FP8 implementation that runs on a single RTX 4090.

🔗Links:
https://playground.bfl.ai/image/generate
https://huggingface.co/spaces/black-forest-labs/FLUX.2-dev

🏷️Tags:
#ai #text2image #flux

4,080

SmartSoftHub

2026 ж. 11 нау., 01:39

📷 Photo

💥 Found a tool for automating routine tasks

🛠️ Fara — a local agent from Microsoft that performs routine tasks instead of the user

🔰 It can analyze files, book tickets, search for products on marketplaces, help with job searches, and more.

🔰 The neural network works locally and your data will not leak anywhere.

🔗Links:
https://github.com/microsoft/fara

🏷️Tags:
#ai #agents #microsoft

3,230

SmartSoftHub

2026 ж. 11 нау., 01:39

📷 Photo

💥 Removing censorship from text neural networks

🛠️ A free project has been released on GitHub that will turn the gentlest LLM into a crazy psychopath ready to provide any instruction

🔰Censorship drops to almost zero: tested on the modest Gemma 3 from Google — the refusal rate dropped from 97% to 3%

🔹 The program removes censorship at the internal settings level;
◦ All parameters are selected automatically;
🔹 Works with all types of models;
🔹 One downside: at least a mid-range computer is needed.

🔗Links:
https://github.com/p-e-w/heretic

🏷️Tags:
#ai #chatgpt #llm

3,050

SmartSoftHub

2026 ж. 11 нау., 01:39

📷 Photo

🌆 RMBG

🛠️ RMBG is a new AI background removal model that significantly outperforms its counterparts.

🔰 It effectively separates the foreground and background for various types of images, providing high accuracy and commercial applicability.

🔗Links:
https://github.com/Bria-AI/RMBG-2.0/

🏷️Tags:
#ai #imageeditor

3,450

SmartSoftHub

2026 ж. 11 нау., 01:39

📷 Photo

💥Local Dream

🛠️ Run Stable Diffusion on Android Devices with Snapdragon NPU acceleration. Also supports CPU/GPU inference.

✨ Features
🎨 txt2img - Generate images from text descriptions
🖼️ img2img - Transform existing images
🎭 inpaint - Redraw selected areas of images

🔗Links:
https://github.com/xororz/local-dream
📲 https://play.google.com/store/apps/details?id=io.github.xororz.localdream

2,520

SmartSoftHub

2026 ж. 11 нау., 01:39

📷 Photo

👍 Found a collection of the best neural networks

🛠️ The service has more than 200 AIs for any copywriting tasks

🔹 Inside there are tools for generating business plans, theses, SEO optimization, e-mail letters, rewriting, and +195 others;
🔹 To make the bot answer in Russian, just add "strictly answer in Russian" to the request;
🔹 The database is free, no registration and no limits.

🔗Links:
https://writify.ai/

🏷️Tags:
#ai #tools

2,590

SmartSoftHub

2026 ж. 11 нау., 01:39

📷 Photo

📥Borgmatic

🛠️ Borgmatic is a powerful backup management software designed for servers and workstations.

🔰It offers client-side encryption and database backup support.

🔰The program is built on Borg Backup and provides flexible configuration options.

🔗Links:
https://github.com/borgmatic-collective/borgmatic/

🏷️Tags:
#backup

2,650

SmartSoftHub

2026 ж. 11 нау., 01:39

🎥 Video

💥Five neural networks integrated into VSCode

🛠️These neural networks write, correct, and explain all code for free, and if you're not satisfied with the results, you can immediately select a different neural network.

⚠️ How to enable AI in VSCode:

1. Create an NVIDIA AI account https://www.nvidia.com/en-eu/.
2. Click on any model and click "Get API Key."
3. Download the CodeGPT extension in https://marketplace.visualstudio.com/items?itemName=DanielSanMedium.dscodegpt.
4. Select the NVIDIA provider and paste the copied API key.
5. You're all set—Llama-3, Mixtral, Gemma, Phi-3, and others are available in the development environment.

🏷️Tags:
#News #VSCode #Dev #AI

2,390

SmartSoftHub

2026 ж. 26 ақп., 07:04

🎥 Video

🗣️ Clone any voice in a couple of seconds

🛠️ A free AI service Vocloner, has been released that will help you voice your text

🔰It's simple: upload a voice file and enter the text you want to voice, then Vocloner will synthesize the text with the cloned voice in a couple of minutes

🔹 Works with short recordings — 10 seconds is enough for cloning;
🔹 Supports all languages;
🔹Perfectly handles phrases of 1-2 sentences;
🔹 Works directly in the browser — no installation needed.

🔗Links:
https://vocloner.com/

🏷️Tags:
#ai #voiceclone

2,370

SmartSoftHub

2026 ж. 21 ақп., 21:01

🤩Tencent has open-sourced HunyuanImage 3.0

🛠️This open text-to-image model has over 80 billion parameters, of which 13 billion are actively used during inference.🔍

🔰The model is comparable to leading closed counterparts and is based on a multimodal language model, fine-tuned for image generation.

🔰It was trained on 5 billion image-text pairs.

🔰The current release focuses on creating images from text descriptions, providing maximum detail.🎨

🔰Future updates promise to add image editing features and multifunctional interaction.

🔰The AI is available on GitHub, HuggingFace, and the model can be tested on Tencent's website.

🔗Links:
https://github.com/Tencent-Hunyuan/HunyuanImage-3.0

https://hunyuan.tencent.com/image/en

🏷️Tags:
#ai #tencent #hunyuan

2,180

SmartSoftHub

2026 ж. 8 ақп., 07:30

😎Alibaba releases new AI models

🛠️A new image editor Qwen-Image-Edit-2509 has been added to the Qwen chatbot.

🔰The AI can preserve the "appearance" of characters during edits and combine several images into one, which reminded users of solutions like Nano Banana from Google.

🔹Alibaba is also releasing Qwen3-Omni — a native multi-modal model that works with text, images, audio, and video and can provide streaming generation of both text and live speech in real time.

🔹The model supports multilingualism (understands text in >100 languages, recognizes speech in many languages, and generates speech in several), and is optimized for low latency in multimodal scenarios.

🔹Both updates strengthen the Qwen ecosystem and bring it closer to practical scenarios in production and interactive applications.

🔰The AI is available directly in Qwen Chat.

🔗Links:
https://chat.qwen.ai/

🏷️Tags:
#alibaba #ai #qwen

1,670

SmartSoftHub

2026 ж. 4 ақп., 12:31

💥Alibaba has open-sourced Wan-Animate — a universal system for animation and character replacement.

🛠️Wan-Animate allows you to bring any image to life based on motion and facial expressions from a reference video, creating realistic clips.🎭

🔰Additionally, the model can completely replace a character in a video, adapting to the lighting and color scheme of the scene, making the integration as natural as possible.✨

🔰In terms of quality, Wan-Animate is already being compared to closed commercial products like Runway and DreamActor, but now it is available to everyone for free in an open format free of charge.🚀

🔗Links:
https://humanaigc.github.io/wan-animate/
https://huggingface.co/spaces/Wan-AI/Wan2.2-Animate

🏷️Tags:
#ai #video

1,460

SmartSoftHub

2025 ж. 6 жел., 10:56

🗣️ Clone any voices and sounds in a couple of minutes

🛠️Real-Time Voice Cloning — a tool that allows you to change your voice in real time, right during a conversation

🔰 It copies all kinds of sounds and can create a whole deepfake speech

🔹 Written in Python and made as simple as possible;
🔹 Just upload a short snippet of voice or sound and get an exact copy that can be used to generate speech;
🔹 Can change voice parameters: pitch, volume, spectrum, and others;
🔹 Works only locally.

🔗Links:
https://github.com/CorentinJ/Real-Time-Voice-Cloning?tab=readme-ov-file

🏷️Tags:
?q=%23clone ?q=%23voice ?q=%23python

1,410

SmartSoftHub

2025 ж. 6 жел., 10:56

😎 Neural network for finding users by description

🛠️Lessie is a free AI-based agent that helps find anyone by description

🔰 It allows users to perform up to 1,500 influencer searches per month
🔹 Collects information from hundreds of open sources: news sites, YouTube, TikTok, Instagram, and many other social networks;
🔹 Can correlate geodata: finds people even in very remote locations;
🔹 Provides complete info: full name, workplace, contacts, and last online time;
🔹 Officially the tool is designed for finding bloggers, clients, investors, and experts, but its functionality is not limited to this;
🔹 Collects results in a table — it will contain hundreds of matches;

🔗Links:
https://lessie.ai/

🏷️Tags:
?q=%23ai ?q=%23osint ?q=%23agent

1,030

SmartSoftHub

2025 ж. 6 жел., 10:56

🎬 Create videos from PDF files

http://%F0%9F%94%B0PDFtoVideo.ai/ is a free online tool for quickly converting PDF files into professional videos with animation and narration

🔹 Understands any text and context, does not distort the meaning, and converts the document into a video sequence;
🔹 Converts up to 80 pages at a time — you can get a whole MOVIE;
🔹 Supports multiple languages

🔗Links:
https://pdftovideo.ai/

🏷Tags:
?q=%23pdf ?q=%23convert ?q=%23video

1,510

SmartSoftHub

2025 ж. 6 жел., 10:56

🗣️MiniMax

🛠MiniMax — a service for generating voiceovers with realistic voices.

🔰It supports different languages, allows uploading texts from files, choosing emotions, and using the functionality for free.

🔗Links:
https://www.minimax.io/audio

🏷Tags:
?q=%23ai ?q=%23voiceover

1,480

SmartSoftHub

2025 ж. 6 жел., 10:56

🪄Magic Canva

🛠Magic Canva is an open multimodal creative assistant that focuses on privacy and local use, replacing tools like Canva and Manus.

🔰It allows users to bring ideas to life with AI by generating images and videos from simple prompts.

🔰Additionally, the Magic Video feature enables the creation of viral short films based on just one sentence.

🔗Links:
https://github.com/11cafe/jaaz

🏷Tags:
?q=%23ai ?q=%23video ?q=%23canva ?q=%23manus

1,880

SmartSoftHub

2025 ж. 6 жел., 10:56

🤩OmniAvatar

🛠️OmniAvatar — a free tool for lip-sync voiceover and animating pictures.
🔰Upload an image and an audio track, choose the quality and description of the voiceover style.

🔗Links:
https://huggingface.co/spaces/jbilcke-hf/OmniAvatar

🏷️Tags:
?q=%23avatar ?q=%23lipsync ?q=%23ai

1,590

SmartSoftHub

2025 ж. 6 жел., 10:56

👨‍💻Wren AI

🛠Wren AI is an AI agent that allows users to request data in natural language and receive accurate SQL queries, charts, and analytical reports within seconds.

🔰It provides capabilities for data interaction, insight generation, and includes a semantic layer that ensures the accuracy of conclusions.

🔰API integration enables the creation of custom agents and chatbots.

🔰The platform supports multiple databases and is compatible with various LLM models.

🔗Links:
https://github.com/Canner/WrenAI

🏷Tags:
?q=%23ai ?q=%23agents ?q=%23sql ?q=%23llm

1,440

SmartSoftHub

2025 ж. 6 жел., 10:56

🪄Activepieces

🛠Activepieces is an open-source solution for AI automation.

🔰The platform offers an intuitive interface suitable for both technical and non-technical users.

🔰Activepieces comes with more than 280 available components.

🔰Users have the option to self-host the application, which ensures data control and enhances security.

🔗Links:
https://github.com/activepieces/activepieces

🏷Tags:
?q=%23ai ?q=%23automation ?q=%23app

1,520

SmartSoftHub

2025 ж. 29 қар., 13:52

🚀Google updated the "AI Photoshop" Gemini

🛠The Gemini 2.5 Flash model from Google can now natively work with images — not only describing them but also creating and editing.

🔰The new model was tested for several weeks on LMArena under the codename nano banana. As a result, the model took first place in the ranking based on blind tests, surpassing GPT Images and Flux Kontext.

🔰Features:

🔹Preserves details, does not distort the image, and does not have the "neural network yellow filter" like GPT Images.
🔹Can combine objects from several photos into one, generate new angles, or change the scene setting.
🔹Step-by-step editing: you can add edits sequentially to achieve the desired result.

👍The model can be tried for free at Google AI Studio.

🔗Links:
https://aistudio.google.com/prompts/new_chat

🏷Tags:
?q=%23gemini ?q=%23ai ?q=%23news

1,390

SmartSoftHub

2025 ж. 29 қар., 13:52

👍 Found one of the best resources for learning Data Science and Machine Learning

🔰Kaggle — offers excellent interactive free courses for DS and ML engineers, upon completion of which you receive a certificate.

🔹 Each course takes only 3–5 hours, includes exercises and additional materials .

🔗Links:
https://www.kaggle.com/learn

🏷Tags:
?q=%23courses ?q=%23mashinelearning ?q=%23python

1,290

SmartSoftHub

2025 ж. 7 қар., 01:44

📱Mobile-use

🛠Mobile-use is an open AI agent for Android and iOS that allows task management using natural language.

🔰Key features include native language control, intelligent navigation through app interfaces, data extraction, and customization capabilities using various language models.

🔗Links:
https://github.com/minitap-ai/mobile-use

🏷Tags:
?q=%23android ?q=%23ios ?q=%23ai ?q=%23agent

1,270

SmartSoftHub

2025 ж. 7 қар., 01:44

💥In China are creating a new dLLM

🛠️ByteDance has announced an experimental language model Seed Diffusion Preview, based on discrete diffusion, which demonstrates an impressive generation speed — 2146 tokens per second, which is 5,4 times faster than similar solutions.🚀

🔰The quality of the generated code meets key benchmarks, opening new horizons for the creation of language models.

🔰Seed Diffusion Preview is already available for testing! 🤖

🔰Previously, their dLLMs were presented by Apple and InceptionLabs

🔗Links:
https://studio.seed.ai/exp/seed_diffusion/

🏷️Tags:
?q=%23llm ?q=%23ai

1,940

SmartSoftHub

2025 ж. 7 қар., 01:44

👍 Found a powerful collection of materials on optimizing agent systems for you

🔰 The repository presents a collection of works on self-developing agents: a ton of info that will make your agents more effective🚀

🔗Links:
https://github.com/EvoAgentX/Awesome-Self-Evolving-Agents?tab=readme-ov-file

🏷Tags:
?q=%23ai ?q=%23agents

1,460

SmartSoftHub

2025 ж. 7 қар., 01:44

💥 We found a database of over 1000 interactive labs and tutorials for you

🚀LabEx is a hands-on learning platform specializing in coding and technology skills

🔰You can develop skills in Linux, DevOps, cybersecurity, programming, data science, and much more on it

🔰The platform offers interactive, AI-guided, structured learning with free labs and real projects.

🔗Links:
https://labex.io/

🏷Tags:
?q=%23skill ?q=%23tutorial ?q=%23devops

1,530

SmartSoftHub

2025 ж. 29 қаз., 02:04

👍 An official prompt guide for GPT-5 has been released

🛠️OpenAI clearly laid out how to get the most out of the new model

🔰 Inside there are ready-made examples, tips, and instructions for business, coding, and everyday tasks

💥We reveal all the capabilities of GPT-5 — here.

🔗Links:
https://cookbook.openai.com/

🏷️Tags:
?q=%23ai ?q=%23prompts

1,520

Showing 29 of 29 posts

No more posts

Loading posts...

Рейтинг

Кіру қажет

Пайдаланушы пікірлері (0)

Әлі пікір жоқ. Өз пікіріңізбен бөлісетін бірінші болыңыз!

Loading reviews...

SmartSoftHub

Join @smartapphub for exclusive from and software content and discussions in 1

Рейтинг

Жалпы рейтинг

#3998Өзгеріс жоқ

Тіл бойынша рейтинг

#876Өзгеріс жоқ

Санат бойынша рейтинг

#30Өзгеріс жоқ

Жазылушылардың өсуі (Соңғы 30 күн)

Барлығы: 5.9K

24 сағаттағы өсу: -25 0%

Соңғы посттар

SmartSoftHub

2026 ж. 11 нау., 01:39

📷 Photo

🏞️ IC-Custom

🛠️IC-Custom is an open-source tool for various image customization based on in-context learning.

🔰It supports the accurate placement of an object from a reference image using a mask or segmentation, and the generation of a new scene with the subject from the reference based on a text description.

🔰IC-Custom can be applied to tasks such as virtual fitting, product placement, furniture arrangement, and character creation.

🔗Links:
https://github.com/TencentARC/IC-Custom/

🏷️Tags:
#ComfyUI #imageeditor #ai