Suche
Live
Bring state-of-the-art agentic skills to the edge with Gemma 4●Supporting Google Account username change in your app●Developer’s Guide to Building ADK Agents with Skills●ADK Go 1.0 Arrives!●Boost Training Goodput: How Continuous Checkpointing Optimizes Reliability in Orbax and Ma...●Announcing ADK for Java 1.0.0: Building the Future of AI Agents in Java●Closing the knowledge gap with agent skills●Jump to play: Building with Gemini & MediaPipe●Build a smart financial assistant with LlamaParse and Gemini 3.1●Developer’s Guide to AI Agent Protocols●Announcing the Colab MCP Server: Connect Any AI Agent to Google Colab●Plan mode is now available in Gemini CLI●Introducing Finish Changes and Outlines, now available in Gemini Code Assist extensions on...●Unleash Your Development Superpowers: Refining the Core Coding Experience●Introducing Wednesday Build Hour●What's new in TensorFlow 2.21●You can't stream the energy: A developer's guide to Google Cloud Next '26 in Vegas●How we built the Google I/O 2026 Save the Date experience●Supercharge your AI agents: The New ADK Integrations Ecosystem●On-Device Function Calling in Google AI Edge Gallery●TorchTPU: Running PyTorch Natively on TPUs at Google Scale●Get ready for Google I/O: Livestream schedule revealed●New enhancements for merchant initiated transactions with the Google Pay API●Build Better AI Agents: 5 Developer Tips from the Agent Bake-Off●Building with Gemini Embedding 2: Agentic multimodal RAG and beyond●Production-Ready AI Agents: 5 Lessons from Refactoring a Monolith●Subagents have arrived in Gemini CLI●Build Long-running AI agents that pause, resume, and never lose context with ADK●MaxText Expands Post-Training Capabilities: Introducing SFT and RL on Single-Host TPUs●Agents CLI in Agent Platform: create to production in one CLI●Bring state-of-the-art agentic skills to the edge with Gemma 4●Supporting Google Account username change in your app●Developer’s Guide to Building ADK Agents with Skills●ADK Go 1.0 Arrives!●Boost Training Goodput: How Continuous Checkpointing Optimizes Reliability in Orbax and Ma...●Announcing ADK for Java 1.0.0: Building the Future of AI Agents in Java●Closing the knowledge gap with agent skills●Jump to play: Building with Gemini & MediaPipe●Build a smart financial assistant with LlamaParse and Gemini 3.1●Developer’s Guide to AI Agent Protocols●Announcing the Colab MCP Server: Connect Any AI Agent to Google Colab●Plan mode is now available in Gemini CLI●Introducing Finish Changes and Outlines, now available in Gemini Code Assist extensions on...●Unleash Your Development Superpowers: Refining the Core Coding Experience●Introducing Wednesday Build Hour●What's new in TensorFlow 2.21●You can't stream the energy: A developer's guide to Google Cloud Next '26 in Vegas●How we built the Google I/O 2026 Save the Date experience●Supercharge your AI agents: The New ADK Integrations Ecosystem●On-Device Function Calling in Google AI Edge Gallery●TorchTPU: Running PyTorch Natively on TPUs at Google Scale●Get ready for Google I/O: Livestream schedule revealed●New enhancements for merchant initiated transactions with the Google Pay API●Build Better AI Agents: 5 Developer Tips from the Agent Bake-Off●Building with Gemini Embedding 2: Agentic multimodal RAG and beyond●Production-Ready AI Agents: 5 Lessons from Refactoring a Monolith●Subagents have arrived in Gemini CLI●Build Long-running AI agents that pause, resume, and never lose context with ADK●MaxText Expands Post-Training Capabilities: Introducing SFT and RL on Single-Host TPUs●Agents CLI in Agent Platform: create to production in one CLI●

KI Force — Die deutsche KI-Wissensreferenz

18 Anbieter · 123+ Quellen · 87591+ Artikel

Enter ↵
★ Aktuell Alle News →
★ Aktuell

Bring state-of-the-art agentic skills to the edge with Gemma 4

Google DeepMind has launched Gemma 4, a family of state-of-the-art open models designed to enable multi-step planning and autonomous agentic workflows directly on-device. The release includes the Google AI Edge Gallery for experimenting with "Agent Skills" and the LiteRT-LM libra...

• google
Neueste Artikel Mehr ansehen →
• developers

ADK Go 1.0 Arrives!

The launch of Agent Development Kit (ADK) for Go 1.0 marks a significant shift from experimental AI scripts to productio...

▶ Tages-Digest — 27. Juni 2026 131 Artikel, 13 relevant
# KI-Tagesüberblick 2026-06-27

Highlights

1. OpenAI stellt GPT-5.6 Sol vor — OpenAI hat sein neues Flaggschiffmodell [GPT-5.6 Sol angekündigt](https://openai.com/index/previewing-gpt-5-6-sol), das Anthropics Claude Mythos 5 in Coding-Benchmarks übertrifft. Der Zugang bleibt vorerst auf ausgewählte Partner beschränkt – eine Folge von Druck durch die US-Regierung, den OpenAI [kritisch kommentiert](https://the-decoder.de/openai-stellt-gpt-5-6-sol-vor-und-uebertrifft-claude-mythos-in-zentralen-benchmarks/).

2. Anthropic bringt Mythos 5 zurück — Nach zwei Wochen Verhandlungen mit der Trump-Administration ist [Anthropics Mythos 5 nun wieder verfügbar](https://www.theverge.com/ai-artificial-intelligence/958458/anthropic-mythos-5-is-back-trump-negotiations), nachdem die Kontrollen gelockert wurden.

3. Intelligentes Model Routing für Coding-Agenten — [Ein neuer Router](https://github.com/workweave/router) leitet Anfragen intelligent an das beste verfügbare Modell weiter und senkt damit die Kosten für Teams, die mit Claude Code, Cursor und anderen Tools arbeiten.

4. Chinesische KI-Modelle schließen Lücke — [Laut New York Times](https://www.nytimes.com/2026/06/25/technology/zai-china-artificial-intelligence-models.html) schließen chinesische KI-Modelle den Leistungsabstand zu Anthropic und OpenAI.

## Modell-Updates

- Claude Code v2.1.195 — [Neue Version](https://github.com/anthropics/claude-code/releases/tag/v2.1.195) mit verbessertem Hook-Matching für hyphenisierte Identifikatoren und Fixes für Voice Dictation auf macOS.
- Claude Agent SDK TypeScript v0.3.195 — [Update](https://github.com/anthropics/claude-agent-sdk-typescript/releases/tag/v0.3.195) mit Query-Reinitialisierung und Bugfixes für Event-Emission.

## Tool-Releases

- Ollama v0.30.11 — [Version](https://github.com/ollama/ollama/releases/tag/v0.30.11) mit automatischer Claude Code-Installation und Thinking-Capability-Detection.
- CrewAI v1.15.1a1 — [Alpha-Release](https://github.com/crewAIInc/crewAI/releases/tag/1.15.1a1) mit verbessertem Telemetrie-Tracking und SSRF-Sicherheitsfix.

## Forschung & Entwicklung

- Self-Improving Agent Skills mit cognee und n8n — [Blog-Artikel](https://blog.n8n.io/skill-loop/) zur Integration von selbstlernenden Agent-Fähigkeiten in Workflows.
▶ Top Playbooks Alle Playbooks →
security

KI-generierten Code kritisch überprüfen und verifizieren

Entwickeln Sie die Gewohnheit, jeden von KI generierten Code sorgfältig zu überprüfen, zu testen und...

beginner
coding

Lerne durch Bauen von Projekten (Aktives Lernen)

Vermeide die 'Tutorial Hell', indem du das Gelernte sofort in die Praxis umsetzt. Baue eigene Projek...

intermediate
Bereiche
🤖

LLM-Anbieter

Alle Modelle weltweit
⚡

AI Tools

Coding, Creative, Productivity
🎦

Videos

YouTube & Tutorials
🔬

Playbooks

Anleitungen zum Umsetzen
📖

Forschung

arXiv Papers & Benchmarks
🇩🇪

DACH

Deutsche KI-Szene