AI Intel

DigitalOcean has announced first-party integration support for OpenAI’s Agents API, allowing developers to run agent applications on their Managed Ag…

DeepSeek-V4-Pro API is currently available with a 75% discount until May 2026, and offers integration updates including a 1M context window for its m…

The GLM-5.3-Flash (ox-alpha) model can now be run locally in 3-bit quantization on 128GB RAM via Unsloth GGUF, showing competitive performance agains…

A high-performance 125B Qwen3.8-Flash Mixture-of-Experts (MoE) model can now be run locally on just 75GB RAM, notably outperforming Claude Opus 4.6.

Claude Code version 2.1.269 has been released, adding a new plugin evaluation feature to run eval suites against Claude Code and obtain reproducible…

Gemini 2.5 Pro and Gemini 2.5 Flash models are now Generally Available, and a preview of agent mode for Gemini Code Assist in VS Code has been releas…

Google Gemini has integrated Lyria 3.5, its most advanced music generation model, offering users the ability to create music with richer arrangements…

A report from Arena.ai on the top 15 models in Agent Arena reveals that Kimi K3 (Max) is the most cost-effective among the top contenders, with a med…

Agent Arena has launched a Pareto frontier analysis tool, allowing users to compare AI model performance on real-world agentic tasks against their me…

A new UI tool is highlighted for simplifying local AI deployment and management by wrapping llama.cpp's complexities into a clean interface, making i…