🚨 AI News | TestingCatalog

Codex got a new "activity" sidebar in the latest update! The button works as a toggle to switch between "Projects & Recent chats" and "Activity" views.

> See chats that are unread, active or awaiting response.

THINKING MACHINES 🔥: A new open-weight model has been released!

Inkling-Small has 276B total parameters, 12B active, and is now available on Huggingface and Tinker Playground.

And it is quite fast 👀

OPENAI 🔥: GPT-5.6 Luna prices got reduced by 80% along with a 20% cut for GPT-5.6 Terra!

GPT-5.6 Sol got a new faster option on the API with a 2.5x speed boost at 2x price.

Many models got overshadowed 👀

Google introduced Gemini Robotics ER 2, a new embodied reasoning model!

Benchmarks 👀

> Success/failure detection: Now operates on raw video feeds rather than static snapshots to catch mid-execution failures like spills, slips, or misalignments.

> General instrument reading: Extends beyond circular dials and sight glasses to include digital displays, linear scales, rulers, and liquid thermometers. We tested it across 10 different types of instruments.

> Enhanced spatial VQA: Improves Visual Question Answering throughGemini’s advancements in multi-modal understanding.

GOOGLE 🔥: Gemini desktop app for macOS now supports "Speak to Window" feature!

> "Today, we’re introducing a new way to create, edit, and summarize with Gemini using just your voice, right where you are already working."

Revolut 🤝 OpenAI

Revolut introduced ChatGPT Go subscription plan as a benefit for their customers.

> Starting from 3 months free on Standard, you can unlock up to 12 months of ChatGPT Go included at no extra cost, depending on your plan.

ChatGPT user base is about to grow quite a lot soon!

CURSOR 🔥: The latest version of Cursor for iOS is now compatible with iPad! Besides that, the app got a new Inbox and a complete PR review experience.

I need more devices for testing 👀

GOOGLE 🔥: Lyria 3.5 has been released on Google Flow Music! Besides that, Flow Music now has covers, lip-sync videos, and an iOS app.

> Meet Lyria 3.5. Experience dynamic vocals, richer musicality, and advanced creative controls with our new flagship model.

> Reimagine your music with Covers. Transform songs into a completely new style while keeping the original structure intact.

> Take the studio with you. Download the Google Flow Music iOS app to create, listen, and share from anywhere.

> Direct lip-synced music videos. Use Gemini Omni Flash and new lip-syncing capabilities to create stunning visuals.

SpaceXAI launches Grok Voice Think Fast 2.0 on Agent Builder

xAI launched Grok Voice Think Fast 2.0, a speech model for voice agents with higher benchmark scores, faster first audio, stronger transcription across 24 languages, and lower latency. It costs $0.08 per audio minute.

🗞 #spacexai @testingcatalog

Bagel Labs launches WorldDiT world model for robotics

Bagel Labs released WorldDiT, an open robotics world model that jointly predicts robot actions and future scenes with one diffusion backbone, delivering strong LIBERO performance at sub-billion scale for on-robot deployment.

🗞 #ai @testingcatalog

GOOGLE 🔥: A new "App" artifact is being developed on Gemini Notebook!

> "Generate an interactive HTML app based on your sources"

Users will be able to customize it via a prompt; examples include interactive dashboards, apps, and games! Imagine playing a game based on books you have added as sources?!

Besides that, Google continues to work on AI Notes and may also add an option to enable or disable watermarking for generated artifacts.

Google is working on interactive Apps for Gemini Notebook

Google is testing an App artifact for Gemini Notebook that could turn source material into runnable apps from prompts. The feature is not live, but aligns with an earlier secure code sandbox release.

🗞 #notebooklm @testingcatalog

Cursor launches Start plan in India for ₹649 per month

Cursor Start launches in India at ₹649/month with tax included, adding Grok 4.5, Composer, more agent requests, and UPI support. The tier targets daily development across desktop, web, iOS, and CLI.

🗞 #cursor @testingcatalog

GOOGLE 🔥: Gemini Notebook is about to get even smarter! In addition, users will be able to control whether watermarks are added to generated artifacts.

"Gemini Notebook just got smarter. Try asking to find new sources from the web"

OPENAI 🔥: New GPT-Live-Transcribe and GPT-Transcribe models are now available on APIs!

> GPT-Live-Transcribe - low-latency live transcription.

> GPT-Transcribe - asynchronous transcription of completed audio files and batch workloads.

OpenAI and researchers from other top companies requesting US government to develop governance measures to be able to “pace AI advancements”

“The world's leading AI companies believe they could be close to automating AI research.”

> We believe that, at some point in the future, AI acceleration for frontier model development may be so high that the world will need to pace the rate of AI advancement.

> We hope to contribute to work led by the U.S. government, alongside other labs and the open-source community, to develop the tools and mechanisms that could make that possible.

> It is hard to predict exactly how much this will accelerate AI progress, but there is a real risk that capability development rapidly accelerates beyond our ability to understand or control the resulting systems.