Comet can now sync across different devices, including mobile and desktop versions.

Claude Opus 4.5 scores 80% on SWE-Bench Verified

Anthropic has released Claude Opus 4.5, its most advanced AI model, delivering higher accuracy in coding, reasoning, and vision tasks. It offers lower costs, improved alignment, and expanded features for developers and enterprise users.

🗞 #claude

Claude Opus 4.5 is rolling out to MAX subscribers on Perplexity. Both "thinking" and "non-thinking" versions.

Windsurf is testing a new model called "Penguin Alpha", potentially from Cognition.

SWE-2? 👀
h/t @ccg_tob2412

OpenAI adds Shopping Research mode to ChatGPT for all users

OpenAI has launched a shopping research mode in ChatGPT, allowing users to get tailored product guidance through real-time data and follow-up prompts. It supports deep research, product comparisons, and integrated checkout via partner merchants.

🗞 #chatgpt

Perplexity is launching PayPal-powered, personalised shopping recommendations in the US, with Instant Checkout!

Black AI Friday soon 👀

BREAKING 🚨: Black Forest Labs released FLUX.2 [pro] and FLUX.2 [flex] image generation models!

- FLUX.2 [pro]: State-of-the-art quality at maximum speed.
- FLUX.2 [flex]: Maximum precision with complete creative control.

Grok 4.1 Thinking claims the top 2 spot on the LMArena text leaderboard (preliminary), with Grok 4.1 following with the top 3.

Top 1 by Jan? 👀

ChatGPT Voice Mode is now integrated directly into the chat! In this way, users can see all messages and other inline widgets during a voice mode conversation.

Rolling out to all users 👀

OpenAI might be preparing ChatGPT Pulse for a broader release to other Plans. Currently, this feature is only available to Pro users.

New Pro-only features soon? 👀

OpenAI to let devs publish their Apps and Workflows to ChatGPT

OpenAI is advancing its developer platform with tools for publishing agent workflows to ChatGPT, integrating image generation, and managing applications via a new dashboard, aligning with upcoming product updates and broader ecosystem access.

🗞 #chatgpt

BREAKING 🚨: OpenAI prepares to enable developers to publish their Apps and Workflows on ChatGPT via the OpenAI Platform. A new Apps directory (GPT Store v2?) will be added to ChatGPT as well.

Other additions:
- "Hermes" feature will let builders publish (and share?) workflows on ChatGPT
- A new Apps dashboard for App creation will be added
- Image generation tool will be added to Agent Builder

Google started rolling out interactive images on Gemini, where users will be able to explore additional details after clicking on certain image elements.

UI from the future 👀

Did you get it already?

Google has extended 🍌Nano Banana Pro to all Paid tiers! Nano Banana Pro 2K and 4K image resolution support will be added in the coming weeks.

Google is rolling out interactive images for learning in Gemini

Google introduced interactive images in the Gemini app, allowing users to click on parts of educational diagrams for definitions and explanations. This AI-driven feature supports academic exploration in subjects like science and biology.

🗞 #gemini

Perplexity launches AI shopping assistant for US users with PayPal

Perplexity is launching a conversational shopping feature for US users, combining search, discovery, and checkout in one thread. Integrated with PayPal, it supports in-chat purchases and retains context across user queries. Mobile rollout is coming soon.

🗞 #perplexity

Black Forest Labs launches FLUX.2 image models

Black Forest Labs released FLUX.2, a suite of image generation and editing models for professionals, featuring high-resolution output, improved prompt fidelity, and public availability across API, open-weight, and open-source formats.

🗞 #ai