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.
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.
❤6🔥1🤔1
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 👀
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 👀
❤9👍3🦄2🤩1
Media is too big
VIEW IN TELEGRAM
PERPLEXITY 🔥: Spaces got upgraded to Projects, a new type of workspaces for collaboration with Perplexity Computer, powered by a shared file system and self-improving Brain memory!
> Between tasks, Brain reviews the Project's files and sessions and updates what it knows, so each task starts with full context from previous work.
> Between tasks, Brain reviews the Project's files and sessions and updates what it knows, so each task starts with full context from previous work.
❤5👍5
Dreamina Seedance 2.5 is now available on Dreamina AI for paid plans in Southeast Asia, the Middle East, Africa, Europe, and South America.
No US 👀
- Native 30s videos
- A new interactive editing experience
- Long video mode (up to 3 minutes)
- Dreamina AI plugins for Maya and Blender
- Up to 50 multimodal references
- More true-to-life lighting and shadows
No US 👀
- Native 30s videos
- A new interactive editing experience
- Long video mode (up to 3 minutes)
- Dreamina AI plugins for Maya and Blender
- Up to 50 multimodal references
- More true-to-life lighting and shadows
❤9🔥2😴1
MiniMax H3 is now available on HailuoAI & MiniMax APIs!
> All-in-One Reference - Creating Videos Using Text, Image, Audio, and Video Inputs
> Precise editing controls - Improve and iterate by adhering to precise guidelines.
> Built for all creative scenarios - from movies and commercials to games, brands, and e-commerce
> 2K video from $0.081/sec, 768p video coming soon, from $0.047/sec
> All-in-One Reference - Creating Videos Using Text, Image, Audio, and Video Inputs
> Precise editing controls - Improve and iterate by adhering to precise guidelines.
> Built for all creative scenarios - from movies and commercials to games, brands, and e-commerce
> 2K video from $0.081/sec, 768p video coming soon, from $0.047/sec
❤6🔥3😴2
This media is not supported in your browser
VIEW IN TELEGRAM
OPENAI 🔥: The built-in web browser in the ChatGPT app is becoming more mature. Now it supports URL suggestions during typing.
Besides that, the ChatGPT Chrome extension can now reference open tabs, highlight to ask, and more!
I may need to try it as a default 👀
Besides that, the ChatGPT Chrome extension can now reference open tabs, highlight to ask, and more!
I may need to try it as a default 👀
❤6🔥4👍1
No more AI Studio for mobile? 👀
But instead, we will be getting something else! Any guesses on what this could be?
> We’ve decided to take an entirely different approach: one where apps emerge naturally, in the course of your everyday conversations with Gemini.
> We’re partnering with the Gemini app team to make that a reality on mobile and desktop.
But instead, we will be getting something else! Any guesses on what this could be?
> We’ve decided to take an entirely different approach: one where apps emerge naturally, in the course of your everyday conversations with Gemini.
> We’re partnering with the Gemini app team to make that a reality on mobile and desktop.
❤11✍5
Gemini for macOS adds new "Speak to Window" feature
Google is rolling out Gemini voice on macOS for English users worldwide. A long press of Fn enables polished dictation and on-screen reasoning to summarize files, rewrite text, and create or edit images without leaving the current window.
🗞 #google @testingcatalog
Google is rolling out Gemini voice on macOS for English users worldwide. A long press of Fn enables polished dictation and on-screen reasoning to summarize files, rewrite text, and create or edit images without leaving the current window.
🗞 #google @testingcatalog
TestingCatalog AI News
Gemini for macOS adds new "Speak to Window" feature
New Gemini voice tools are rolling out globally in English on macOS, with Fn-key dictation and optional reasoning based on on-screen context.
❤5👍1
Google started rolling out Gemini Spark to Google AI Pro users outside the U.S.
> Spark is your personal AI agent that works in the background 24/7 to get things done under your direction, handling the heavy lifting so you can focus on what matters.
Let’s see how long will that take. The next big upgrade there would be to get support for real MCPs (which is in the works at least).
> Spark is your personal AI agent that works in the background 24/7 to get things done under your direction, handling the heavy lifting so you can focus on what matters.
Let’s see how long will that take. The next big upgrade there would be to get support for real MCPs (which is in the works at least).
👍6❤5
OPENAI 🔥: A new model family named “Astra” has been teased! It is designed to solve long-running tasks where multiple agents work together on the problem.
As per announcement from OpenAI, this model solved 10 significant math problems at a cost of $2000 at Sol API prices.
According to The Information, OpenAI doesn’t have a decision yet if this model will be released as GPT-6 or GPT-5.7.
“Astra” is currently undergoing the Trump testing phase.
As per announcement from OpenAI, this model solved 10 significant math problems at a cost of $2000 at Sol API prices.
According to The Information, OpenAI doesn’t have a decision yet if this model will be released as GPT-6 or GPT-5.7.
“Astra” is currently undergoing the Trump testing phase.
❤8🔥2
This media is not supported in your browser
VIEW IN TELEGRAM
OPENAI 🔥: ChatGPT Pets now support voice mode as well! Users can activate voice mode and approve or stop tasks directly from the pet area.
As it has been foretold 👀
As it has been foretold 👀
❤7👍2
Thinking Machines launched open-weight Inkling-Small
Thinking Machines released open-weight Inkling-Small, a 276B MoE model with 12B active parameters, matching Inkling at one-quarter the size. It supports multimodal chat, 1M-token context, fine-tuning, and strong reasoning and coding results.
🗞 #thinkingmachines @testingcatalog
Thinking Machines released open-weight Inkling-Small, a 276B MoE model with 12B active parameters, matching Inkling at one-quarter the size. It supports multimodal chat, 1M-token context, fine-tuning, and strong reasoning and coding results.
🗞 #thinkingmachines @testingcatalog
TestingCatalog AI News
Thinking Machines launched open-weight Inkling-Small
Inkling-Small is now available with open weights, Tinker fine-tuning and multimodal Playground chat, using 12B active parameters and up to 1M tokens.
❤6👍2
Media is too big
VIEW IN TELEGRAM
SPACEXAI 🔥: Grok Imagine got upgraded with Voice Consistency, Multi-References, Text-to-Video, and native 1080p support!
Users will be able to @-tag reference files via the Grok Imagine agent to reuse previously created characters with their consistent voices.
> Now available for SuperGrok Plus and Heavy subscribers, rolling out to all tiers over the next few days.
Users will be able to @-tag reference files via the Grok Imagine agent to reuse previously created characters with their consistent voices.
> Now available for SuperGrok Plus and Heavy subscribers, rolling out to all tiers over the next few days.
❤8👍2
xAI adds character references and 1080p to Imagine Video 1.5
xAI expanded Imagine Video 1.5 with prompt-only video creation, image and voice references, multi-reference control, and native 1080p output. The update adds tighter control over character, scene, product, and continuity via Grok and API.
🗞 #spacexai @testingcatalog
xAI expanded Imagine Video 1.5 with prompt-only video creation, image and voice references, multi-reference control, and native 1080p output. The update adds tighter control over character, scene, product, and continuity via Grok and API.
🗞 #spacexai @testingcatalog
TestingCatalog AI News
xAI adds character references and 1080p to Imagine Video 1.5
Imagine Video 1.5 adds up to seven image references, voice consistency, prompt-only generation, and native 1080p for Grok users.
❤3👍3
SPACEXAI 🔥: A new SuperGrok Plus subscription plan has been introduced! It comes with significantly higher usage limits and is priced at $100 per month.
> The plan also includes app creation and one-tap deployment support.
> Significantly higher usage across Chat, Imagine, Voice & Build.
> Lightning-fast replies and 1080p videos.
> Priority access at peak times and Early Access to new features.
> The plan also includes app creation and one-tap deployment support.
> Significantly higher usage across Chat, Imagine, Voice & Build.
> Lightning-fast replies and 1080p videos.
> Priority access at peak times and Early Access to new features.
❤4👍2