A practical perspective for individuals
We turn new tools and features into one question: what can you do with them today, without unnecessary hype?
Available now
A specific item, procedure and conditions verified in the original source.
What you can tryOpen-source code and SDK for the Muse Gadgets project, for custom devices with the Muse agent.
HowProgram an ESP32 board or set up a Raspberry Pi using the published SDKs for the Muse Gadgets project.
TermsYour own ESP32 board or Raspberry Pi computer.
Check access at the source ↗What you can tryQwen-Image-2.1 for generating and editing images, including transparent RGBA images
HowOpen the Qwen-Image-2.1 model demo on Hugging Face and try generating or editing an image.
TermsThe research license prohibits commercial use; commercial use requires requesting a separate license from the Qwen team.
Check access at the source ↗What you can tryBootLoops, an open-source tool for precise scientific calculations with language models
HowDownload the source code for BootLoops from GitHub.
Check access at the source ↗What you can trySpeech feature for generating spoken word with optional musical backing
HowOn the Suno service website, open the Create tab, select Speech, and in Simple mode enter a description of the desired recording.
TermsPublic beta; the maximum recording length is approximately eight minutes.
Check access at the source ↗What you can tryREST API and GraphQL API for GitHub Copilot code review with an effort level selection for each request
HowRequest a code review via the REST API or GraphQL API and optionally set the effort level for the given request.
TermsCopilot Pro, Pro+, Max, Business, or Enterprise plan.
Check access at the source ↗What you can tryClef and Clef-flash models for AI agent decision-making over text and images
HowDownload the Clef or Clef-flash model from the Hugging Face platform and try assigning probabilities to predefined options.
TermsApache-2.0 license.
Check access at the source ↗Practical radar
From 30 July, OpenAI reduced the price of GPT-5.6 Luna by 80 % (to 0.20 $/1.20 $ per million tokens) and Terra by 20 %. According to the company, the cuts stem from savings achieved by optimizing infrastructure with Sol; commentators point to pressure from low-cost competitors.
Anyone using OpenAI API for their own projects or tools with a low-cost model (Luna) will pay significantly less for the same work than before, as illustrated by Simon Willison, who switched from Gemini to Luna because of the price.
To accompany the release of the open model Muse Glimmer (30B, Apache 2.0), Meta published an extensive essay by Mark Zuckerberg defending distillation and its open strategy against OpenAI and Anthropic; it also promised to open the weights of Muse Spark 1.2.
New: Mark Zuckerberg published an essay (6000+ words) on AI philosophy, governance and the strategy of Meta; The announcement includes Muse Code, an AI agent for the terminal; Meta strategically positions open-weight models as an alternative to proprietary approaches from OpenAI and Anthropic; The essay criticizes lobbying by competitors and mentions competition from Chinese open-weight models; The new strategy of Meta focuses on differentiation through an open-source approach
Developers and enthusiasts can run a full-fledged agent for coding or working with files or images directly on their Mac or PC, without depending on the cloud or sending data to a third party.
GPT-6 Sol and GPT-6 Luna from OpenAI are now generally available on the Amazon Bedrock platform at lower prices than GPT-5.6; according to OpenAI, GPT-6 Sol makes about half as many factual errors as GPT-5.6 Sol.
Developers working with the Amazon Bedrock platform gain another model choice based on the task type: assign more demanding tasks to GPT-6 Sol and simple, frequent tasks to GPT-6 Luna, and adjust the reasoning effort level as needed.
The computer use feature is in public preview in the GitHub Copilot CLI and GitHub Copilot app tools for macOS and Windows. According to GitHub, it allows control of desktop applications. It requires user consent, and organizations can disable it.
According to GitHub, the tool can be entrusted with tasks such as summarizing browser notifications or editing a presentation; the user decides whether to allow control of a specific application.
The Claude Opus 5.5 model from Anthropic is now available in GitHub Copilot for Pro+, Max, Business and Enterprise plans. According to Anthropic, it handles tasks comparably to Opus 5, but with fewer steps and tokens.
Users on Copilot Pro+, Max, Business and Enterprise plans can select the Claude Opus 5.5 model in the model picker for agentic coding and long-running tasks once the gradual rollout reaches them.
The company Black Forest Labs has released the Flux 3 Image model. According to the company, it handles multi-step edits without changing other parts of the image. It supports up to ten reference images and output of up to 4K. A free demo is available.
When creating images, you can try step-by-step edits and working with up to ten reference images; preserving other parts during edits is, for now, a claim made by the company.
The Ideogram 4.5 model is available on the Ideogram platform and via API. According to Ideogram, it changes only the requested parts of an image. It offers four quality tiers at native 2K resolution for 0.8 to 22 cents per image.
When editing a photo, you can try changing a specific element with the declared preservation of its surroundings. It is advisable to verify the reliability of this feature on your own image.
The automations feature in GitHub Copilot in the VS Code editor allows running agent tasks hourly, daily, weekly, or on demand. A task definition can be created from a template or using a custom prompt. The feature is available in preview.
Part of an overview of multiple AI topics; only this event has been covered.
A developer can prepare their own task definition and run it on demand, without having to build it again each time.
The HydraFusion research preview is now available in Visual Studio Code and the GitHub Copilot app. It coordinates multiple models and supports the Copilot Pro, Pro+, Business, and Enterprise plans. Company access requires an admin to enable preview features.
If you're using a supported plan, you can try HydraFusion directly in Visual Studio Code or the GitHub Copilot app, without switching to the Copilot CLI tool.
The Guided Vision feature is also available in Google TalkBack. On devices running Android 9 and later, an accessibility shortcut can be set up. The audio description of images from the camera helps with reading text and recognizing objects; according to Google, it is not intended for navigation.
New: Guided Vision is also available in Google TalkBack.; On devices running Android 9 and later, an accessibility shortcut can be set up for Guided Vision.; Google warns against using the feature for navigation and obstacle detection.
Blind and low-vision users can launch audio image descriptions directly from the Google TalkBack menu and ask follow-up questions about text or objects captured by the camera.
The ChatGPT product now lets you virtually try on clothing and accessories using your own photo and save selected products using the Favorites feature. OpenAI announced the global availability of both features.
When deciding on clothing or accessories, you can use a preview on your own photo and save selected products to return to later.
The GitHub Copilot app offers optional local sandboxing in public preview, which restricts the agent's access to files, network, and credentials. It is disabled by default and configured per project.
Developers using the GitHub Copilot app can restrict the agent's access to files, network, and credentials for an entire project, or just for the current session using the /sandbox on command.
When personal experience proves useful
Look at the same topics through the lens of processes, risks, and implementation within a company.
AI Radar monitors Czech and international sources every day, looking for changes that truly deserve attention.
The result is a quick overview of what has actually changed in the AI world, rather than another stream of articles.
Each row represents one event — not one article. At a glance, you can see its significance, credibility and main point.
Agent mode is available to all paying users
Until now, the mode was available only on the highest plan; it is now available on all paid tiers without a waitlist.
When you open an event, you will find a more detailed summary, an explanation of its significance, and links to the original sources. Quick updates are available separately in AI Flash; important topics move to Today’s Overview once more information is added.
We first summarize each event in the same way for everyone. Based on those same facts, we then explain what the change means for your own use and what it could mean for how a company operates.
Today’s briefing is the same for everyone. Pages For individuals and For businesses can be found in the main navigation — they select only events relevant to the given use case.
