Staying updated in AI shouldn’t take hours.

@CryptoEternalAI pulled out the updates that actually matter—from Manus, Together AI, Qwen, DeepSeek, Binance, NVIDIA, Anthropic, and more—so you don’t have to.

Here’s your shortcut to staying in the loop.🧵

1/ @ManusAI_HQ shared two slick use cases that make learning more fun and practical.

- FastAPI Course Creator: Build an interactive site to teach coding. Simple and clean.
- Anki Cards Generator: Drag, drop, and instantly turn content into flashcards.

Manus delivers EDUCATION 2.0 🧑‍🏫

1/ FastAPI Course Creation

Manus built a learning website. Interactive. Immersive. All your stuff in one sweet spot online.

Hit RT + throw us your education challenge below. We'll show you how Manus makes it real. https://t.co/3F2xxplxaj

ManusAI

Our mission is to preserve humanity's greatest creation, AI, onchain — ensuring it remains accessible to all, resistant to censorship & beneficial for everyone.

2/ @togethercomputejust launched Together Chat.

Chat with open-source models like DeepSeek R1 & LLaMA 3.3 70B—for free.

Now supports: web search, image generation & analysis, coding, branching chats, and “thinking” mode.

2/ @togethercomputejust launched Together Chat.

Chat with open-source models like DeepSeek R1 & LLaMA 3.3 70B—for free.

Now supports: web search, image generation & analysis, coding, branching chats, and “thinking” mode.

https://x.com/nutlope/status/1904204127602618875

3/ @Alibaba_Qwen dropped a powerful 32B VLM.

- Optimized with reinforcement learning
- Big boosts in math reasoning + human preference
- Apache 2.0 license — great pick for local setups

3/ @Alibaba_Qwen dropped a powerful 32B VLM.

- Optimized with reinforcement learning
- Big boosts in math reasoning + human preference
- Apache 2.0 license — great pick for local setups

https://x.com/Alibaba_Qwen/status/1904227859616641534

4/ @deepseek_ai quietly rolled out DeepSeek-V3-0324 on @huggingface.

- Just 700GB, MIT license
- Competes with Claude 3.7 in coding & math
- Runs at 20 tokens/sec on a Mac Studio

4/ @deepseek_ai quietly rolled out DeepSeek-V3-0324 on @huggingface.

- Just 700GB, MIT license
- Competes with Claude 3.7 in coding & math
- Runs at 20 tokens/sec on a Mac Studio

https://x.com/awnihannun/status/1904177084609827054

5/ Eternal AI is now integrated with @BinanceWallet.

You can access @CryptoEternalAI directly from the Binance app and launch onchain AI agents on @BNBCHAIN—no coding required.

Onchain AI just got a whole lot easier.

5/ Eternal AI is now integrated with @BinanceWallet.

You can access @CryptoEternalAI directly from the Binance app and launch onchain AI agents on @BNBCHAIN—no coding required.

Onchain AI just got a whole lot easier.

https://x.com/CryptoEternalAI/status/1902947109026009138

6/ @NVIDIAAI published MambaVision on Hugging Face.

- A hybrid vision model combining Mamba + Transformers
- Delivers SOTA performance on ImageNet-1K in both accuracy and speed
- Great for image classification, feature extraction, and more

6/ @NVIDIAAI published MambaVision on Hugging Face.

- A hybrid vision model combining Mamba + Transformers
- Delivers SOTA performance on ImageNet-1K in both accuracy and speed
- Great for image classification, feature extraction, and more

https://huggingface.co/nvidia/MambaVision-L3-256-21K

7/ Highlights from NVIDIA GTC 2025

- DGX Spark: Pocket-sized AI beast with 128GB memory 
- DGX Station: 20 petaflops + 72 CPU cores for heavy AI workloads 
- New GPU roadmap includes Blackwell Ultra in 2025, Vera Rubin in 2026–27, and Feynman in 2028.

7/ Highlights from NVIDIA GTC 2025

- DGX Spark: Pocket-sized AI beast with 128GB memory 
- DGX Station: 20 petaflops + 72 CPU cores for heavy AI workloads 
- New GPU roadmap includes Blackwell Ultra in 2025, Vera Rubin in 2026–27, and Feynman in 2028.

https://www.youtube.com/watch?v=_waPvOwL9Z8

8/ @AnthropicAI enabled web search for Claude.

Claude can now browse the internet and pull real-time info—with inline citations so you can verify sources.

A much-needed update. Claude’s now playing catch-up with ChatGPT and Grok.

8/ @AnthropicAI enabled web search for Claude.

Claude can now browse the internet and pull real-time info—with inline citations so you can verify sources.

A much-needed update. Claude’s now playing catch-up with ChatGPT and Grok.

https://x.com/AnthropicAI/status/1902765011727999046

1/ @ManusAI_HQ shared two slick use cases that make learning more fun and practical.

- FastAPI Course Creator: Build an interactive site to teach coding. Simple and clean.
- Anki Cards Generator: Drag, drop, and instantly turn content into flashcards.

https://x.com/ManusAI_HQ/status/1904195897979740288