Core Ai News
2.44K subscribers
1.24K photos
193 videos
1 file
331 links
• Daily AI updates
• Latest breakthroughs
• Machine learning & deep learning
• AGI & LLM news
• AI research & trends

Other channels
- @CorePrompts
- @CoreAgents
- @Coretutorial
- @CoreUtil

X - @Mrxseek

Run by a Human - @SummonMRX
Download Telegram
BREAKING 🚨: OpenAI is expanding ChatGPT Go plan globally at $8 a month with increased usage limits and unlimited access to GPT-5.2 Instant.

Would you get it? 👀
4
Anthropic is working on the Voice Mode support! Operating Claude Cowork agents with voice would be a huge improvement.

It feels like it will work as a toggle in a similar way to how it has been done on ChatGPT.

Would you use it? 👀
This media is not supported in your browser
VIEW IN TELEGRAM
BREAKING 🚨: Anthropic is working on "Knowledge Bases" for Claude Cowork. KBs seem to be a new concept of topic-specific memories, which Claude will automatically manage! And a bunch of other new things.

Internal Instruction 👀

"These are persistent knowledge repositories. Proactively check them for relevant context when answering questions. When you learn new information about a KB's topic (preferences, decisions, facts, lessons learned), add it to the appropriate KB incrementally."
👍1
This media is not supported in your browser
VIEW IN TELEGRAM
This AI agent learns how to use a computer by watching a human do a task, and then it can do similar tasks on its own.

Free & open source
1👍1
This media is not supported in your browser
VIEW IN TELEGRAM
NovaSR is an ultra-tiny 50-kilobyte AI model that enhances audio quality in real-time, upsampling at over 3,500 times real-time speed.

This open-source and includes a free Hugging Face evaluation. 👀

@CoreAti
👍1
VIBE is an open-source image editor that generates 2K resolution images in ~4 seconds on an Nvidia H100. It offers a free Hugging Face demo. 👀

https://riko0.github.io/VIBE/ @CoreAti
This media is not supported in your browser
VIEW IN TELEGRAM
ShapeR, by Meta, reconstructs accurate 3D models from videos/photos, generating individual, metric-accurate 3D objects.

Code and models are available on GitHub. 👀
Pocket TTS is a compact text-to-speech generator (100M parameters) offering real-time, ultra-low latency operation on a CPU with voice cloning for English.

It is open-source with an online demo. 👀


@CoreAti
VerseCrafter, from Tencent, generates videos from single images, providing user control over camera and object movements.

A GitHub repository and Blender add-on are available. 👀

@CoreAti
Top Open-Source AI Innovations This Week!

ShowUI-Aloha (AI agent learns tasks)
NovaSR (tiny audio enhancer)
VIBE (fast image editor)
ShapeR (3D model reconstruction)
Pocket TTS (real-time text-to-speech)
VerseCrafter (video from image)
2🔥1
Media is too big
VIEW IN TELEGRAM
BREAKING 🚨: Notion is working on a big set of new AI features for Notion Agents.

- Custom MCP support
- New agent integrations with Linear and Ramp
- Notion Mail and Notion Calendar triggers for custom agents
- Custom workers (tools) for agents
- Custom Connectors
- New Library and Feed tabs
- AI Co-editor
- And Computer Use for AI agents! 👀
1
Z AI released GLM-4.7-Flash, a new 30B open source model with a 59.2% achievement on SWE Bench Verified.

⚡️ AI
Now you can run many SWE agents like Claude Code, Codex CLI, Gemini CLI and others on remote VMs, managed with a single Agents API released by Blackbox AI.

AI Agents outsource 🤖
2
The full-featured Typeless keyboard is now live on Android, enabling voice dictation across all your apps.

Complete cross-platform 🔥

Link https://android.typeless.com/
🔥1
Google expanded Gemini to 23 more languages.

70+ languages and 230+ countries are supported by now.
A true Gemini 3 Pro (GA?) is being A/B tested on AI Studio.

Robot SVG bench 🤖
1