Google DeepMind releases Gemma 4 12B: a unified encoder-free multimodal model
Gemma 4 12B is Google DeepMind's latest medium-scale multimodal model, featuring an encoder-free unified architecture…
AI news monthly archive: June 2026 features 711 curated AI news items covering model releases, product updates, industry news, papers, tutorials and opinions, each with an AI summary, source credit and original link.
Gemma 4 12B is Google DeepMind's latest medium-scale multimodal model, featuring an encoder-free unified architecture…
Anthropic analyzed 832 malicious accounts banned for policy violations (March 2025 to March 2026) and mapped their…
Well-known AI critic Gary Marcus explores the fundamental challenges facing AI development in his column on…
The Microsoft Superintelligence team released two image models, MAI-Image-2.5 and MAI-Image-2.5-Flash. MAI-Image-2.5…
SenseTime has open-sourced SenseNova-Skills, an AI office skills suite. It is an open-source skill collection designed…
Microsoft AI published an article explaining how its values in health are implemented in products and engineering…
Holo3.1 is a family of computer-use agents built on the Qwen model family, designed to improve robustness in desktop,…
MiniCPM-V 4.6 is now fully supported in vLLM v0.22.0! No custom branch, no extra compilation. Just pull the prebuilt…
Anthropic is expanding its Project Glasswing program to about 150 new organizations, following an initial cohort of…
At the Build 2026 keynote, Microsoft announced seven new MAI models covering image, voice, transcription, reasoning,…
Anthropic last week released dynamic workflows in Claude Code, allowing Claude to write a custom harness for a task on…
Ling-2.6-1T-base is a trillion-parameter MoE base model open-sourced by Ant inclusionAI (about 1T total parameters,…
The Microsoft MAI team released seven new MAI models, which will be made available to developers through OpenRouter,…
Anthropic's interpretability team presented new progress in its Circuits research. To distinguish model features with…
OpenAI's Codex has added features including plugins, sites, and annotations. The new features are designed to help…
SK hynix Chairman Chey Tae-won announced plans to double overall wafer capacity over the next five years to address…
Microsoft released the coding model MAI-Code-1-Flash, trained directly with the production GitHub Copilot harness and…
Thank you for using our models to create these complex charts and diagrams. It's great to see challenging information…
The author created and open-sourced "Clean Junk.skill" based on Codex, which performs read-only scans on Mac and…
The Ant Ling team proposed KPop to address training collapse caused by inconsistent probability distributions between…
OpenAI's GPT-5.5, GPT-5.4, and the Codex coding agent are now generally available on Amazon Bedrock. Developers can…
🎯 Andrej Karpathy on how to learn.
According to OpenRouter platform data, open model usage has grown significantly since 2025. The latest data shows…
As the preferred cloud partner for MiniMax M3, Together AI described the engineering behind its production-grade…
In an interview, Sam Altman said AI should not be designed to pursue goals detached from human needs, and that humans…
M3 on Cloudflare AI Gateway, day one ⚡ Frontier coding capability, 1M context, native multimodal, now callable with a…
Alphabet plans to raise $80 billion by selling stock to support its AI buildout.
Long-form writing needs more space. You can now edit longer articles in full-screen mode and save them to your library…
We just launched the ability to build apps connected to Gmail, Drive, Sheets and more directly inside @GoogleAIStudio,…
Hackers exploited a vulnerability in Meta's AI customer support chatbot to steal high-value Instagram usernames and…
Florida has filed a lawsuit against OpenAI and its CEO Sam Altman. The state attorney general accused Altman of…
From cluttered to clear. Watch parallel sub-agents in @Antigravity categorize and rename hundreds of marketing assets,…
Qwen3.7-Plus deeply integrates vision and language to achieve an end-to-end loop of see, think, write, act, and…
Introducing Search as Code, our new search architecture built for AI agents. It writes Python code directly to call…
OpenAI outlined its position on AI policy and political advocacy, including support for transparency, thoughtful…
AI company Anthropic confidentially submitted an initial public offering (IPO) application on Monday.
Easily add yourself to your Gemini video creations. Here's how to use Gemini Omni to create a digital avatar that…
Claude Code has added dynamic workflows, allowing the model to improvise and coordinate multi-agent frameworks at…
Cognition announced Devin Desktop, positioned as the next generation of Windsurf. It makes the Agent Command Center…
Auto Router now lets you adjust its tradeoff between cost and quality. A new `cost_quality_tradeoff` parameter ranges…
Page 17 of 18
Sponsored
Orange Relay (pro.nat6.net) is a premium AI API relay — an excellent transfer station with competitive pricing, reliable quality and broad model coverage including Claude, GPT and Gemini.
Visit Orange Relay