AI News

Latest news and trends from the world of artificial intelligence

Register or sign in to read AI news translated into Czech.
Introducing GPT-Live
OpenAI 8. 7. 2026
Introducing GPT-Live

OpenAI has launched GPT-Live, a full-duplex voice model for ChatGPT that enables simultaneous listening and speaking, smarter reasoning via background delegation, and more natural conversational flow.

ChatGPT can now listen and talk at the same time, making AI conversations seem more human
The Decoder 8. 7. 2026
ChatGPT can now listen and talk at the same time, making AI conversations seem more human

OpenAI's new GPT-Live voice model uses full-duplex architecture to allow simultaneous speaking and listening, significantly improving conversational naturalness and reasoning capabilities by delegating complex tasks to GPT-5.5.

The Decoder 8. 7. 2026
Mistral enters robotics with Robostral Navigate, an 8B model that steers robots using just one camera

Mistral has unveiled Robostral Navigate, an 8B parameter model for robot navigation that achieves high success rates using only a single RGB camera, trained entirely in simulation.

Anthropic's fix for Fable 5's high cost is turning it into a manager that delegates to Sonnet 5
The Decoder 8. 7. 2026
Anthropic's fix for Fable 5's high cost is turning it into a manager that delegates to Sonnet 5

Anthropic is promoting 'Advisor' and 'Orchestrator' patterns to reduce costs for Fable 5, where the larger model acts as a planner and delegates execution to the more efficient Sonnet 5.

Anthropic's Claude Fable 5 dominates new industry benchmarks at a steep premium
The Decoder 8. 7. 2026
Anthropic's Claude Fable 5 dominates new industry benchmarks at a steep premium

Artificial Analysis has launched six new industry-specific AI capability indices. Anthropic's Claude Fable 5 leads all categories, but at a significantly higher cost compared to open-weights models like GLM-5.2 or DeepSeek V4, sparking debate over the value proposition of frontier models for enterprise tasks.

The Decoder 8. 7. 2026
Google Deepmind adds background execution and MCP support to Gemini API managed agents

Google Deepmind has updated its Managed Agents in the Gemini API with background execution, MCP support, custom function integration, and improved credential management, allowing for more robust agentic workflows.

The Decoder 8. 7. 2026
Chinese AI startup MiniMax plans to open-source a 2.7 trillion parameter model later this year

Chinese AI startup MiniMax is developing a 2.7 trillion parameter model, internally called M3 Pro, which it plans to release as open-source later this year, potentially becoming the largest Chinese AI model to date.

The Decoder 8. 7. 2026
Meta tests always-on AI glasses that capture your entire day

Meta is prototyping AI glasses with a 'Super Sensing' mode that continuously records audio and video without an external indicator light, raising significant privacy concerns while aiming to provide users with a persistent AI memory assistant.

Muse Image is technically impressive, but Meta's use of Instagram photos raises questions
The Decoder 8. 7. 2026
Muse Image is technically impressive, but Meta's use of Instagram photos raises questions

Meta's new agentic image model, Muse Image, offers impressive self-refinement capabilities but faces scrutiny for allowing users to generate images of people using their public Instagram photos without explicit consent, raising potential GDPR and EU AI Act compliance issues.

OpenAI's GPT-5.6 launches Thursday after a delay forced by the U.S. government
The Decoder 8. 7. 2026
OpenAI's GPT-5.6 launches Thursday after a delay forced by the U.S. government

OpenAI has released its GPT-5.6 models following a government-mandated delay. The new models, including the 'Sol' variant, demonstrate superior performance and token efficiency compared to Anthropic's Claude Mythos 5 in coding and cybersecurity benchmarks.

Showing 21 to 30 of 4196 items