TestingCatalog AI News 🗞 – Telegram
TestingCatalog AI News 🗞
4.05K subscribers
2.69K photos
325 videos
40 files
3.78K links
Reporting AI nonsense. A future news media, driven by virtual assistants 🤖
Download Telegram
Kimi-K2 Reasoning model is coming soon 👀
🔥4👍1
This media is not supported in your browser
VIEW IN TELEGRAM
Manus AI now supports slide templates, so you can follow your brand guidelines and styles while generating new slide decks.

First 👀
👍6
Gemini Deep Research can now connect to your Gmail, Docs, Drive and Chat.

"Deep Research will be pulling in information directly from your Drive, including Docs, Slides, Sheets and PDFs"

Testing time 🔥
🔥7👍1
Google connects Gemini Deep Research with Workspace data

Google has introduced Gemini Deep Research, allowing desktop users to combine data from Gmail, Drive, and Chat with web content for integrated research. This update strengthens support for professionals using Google Workspace tools.

🗞 #gemini
3
BREAKING 🚨: Apple will be using a custom 1.2T params model from Google to power Siri and also working on its own 1T params foundational model, as reported by Bloomberg.

A trillion params model club 🤖
👍6🤡3👀1
BREAKING 🚨: Perplexity and Snapchat partner to bring AI search into Snapchat app starting from January 2026.

“Perplexity’s AI-powered answer engine will let Snapchatters ask questions and get clear, conversational answers.”

Snaplexity 👻
👍5
BREAKING 🚨: First beta users are getting early access to Comet browser on Android!
🔥10👎3👍1
Perplexity starts testing Comet browser for Android in beta

Perplexity has begun a limited beta of its Comet browser for Android, mirroring the desktop layout with features like homepage widgets, tab control, dark mode, and ad blocking, while broader functionality and rollout plans remain pending.

🗞 #perplexity
👍3
BREAKING 🚨: GPT-5.1 confirmed as new traces of "gpt-5-1-thinking" have been spotted on ChatGPT.

Gemini 3 vs GPT 5.1 soon 🥊
9👀1
BREAKING 🚨: Kimi Moonshot is preparing to announce "kimi-k2-thinking" and "kimi-k2-thinking-turbo" as these models appear on the API Playground.
👍4
OpenAI readies GPT-5.1 Thinking model ahead of Gemini 3 Pro

Traces of GPT-5.1, including a “Thinking” variant optimized for multi-step reasoning, suggest an upcoming release from OpenAI. Variants may differ by context size and compute limits, with reduced API pricing also rumored.

🗞 #chatgpt
👍3
Who will win on LM Arena in 2025? 🤖

As we approach the last LLM battle of the year between Google and OpenAI, with their upcoming contenders: "Gemini 3 Pro" and "GPT-5.1 Thinking".
Anonymous Poll
74%
Gemini 3 Pro
16%
GPT-5.1 Thinking
10%
Something else 👀
Google Cloud unveils Ironwood TPU and Axion instances

Google Cloud introduces Ironwood TPUs and Axion Arm-based instances targeting AI training and inference at scale. Ironwood offers a 10× gain over TPU v5p, while Axion shows strong price-performance for enterprises shifting workloads.

🗞 #ai
👍2🔥2
Moonshot AI unveils Kimi K2 Thinking with 44.9% achievement on HLE

Moonshot AI's Kimi K2, a trillion-parameter open-weight model, outperforms GPT-4.1 in reasoning and coding with groundbreaking accuracy and efficiency. Its MoE architecture and agentic tools mark a shift toward autonomous, task-oriented AI.

🗞 #kimi
👍4🔥1
ICYMI: Google Flow updates from 11/4/2025

- Adjusting the camera position or motion in a previously generated video is now available to Ultra subscribers.
- Prompt Expanders support with Ingredients to Video.
👍3
Media is too big
VIEW IN TELEGRAM
BREAKING 🚨: Comet browser got a major upgrade for its agentic capabilities.

- It can work longer and on more complex jobs across larger periods of time.
- It performs 23% better than its predecessor.

Longer context was a top request from all the users! Now it can write a weekly recap newsletter for me in one shot as well as seamlessly work with spreadsheets.

Testing time 👀
🔥6👍4
A new stealth model, "Polaris Alpha," is now available on OpenRouter. 256k context window and a quite good performance.

It is worth to test 👀
👍52