AI Intel

An article suggests that the performance and viability of running local AI models have substantially improved, making them a more practical option.

The Hashicorp founder believes local models are not yet "good enough," a sentiment contested by the community which finds SLMs perfectly adequate for…

Claude Code version 2.1.178 introduces new features including `Tool(param:value)` syntax for permission rules to match tool input parameters and the…

llama.cpp has released version b9670, addressing NVFP4 edge cases in llama-graph and implementing necessary post-GEMM MUL for LORA dequantization and…

SpaceX is reportedly acquiring Cursor for $60 billion, indicating a significant investment in potential AI or software capabilities.

OpenAI acknowledges high error rates experienced by some Codex users, attributed to "model at capacity" issues, and is actively working to restore st…

The OpenAI Developers plugin in Codex allows users to accelerate building with OpenAI tools by simplifying API key setup, documentation access, and d…

Grok Build has released an Agent Dashboard, enabling users to manage multiple AI agents simultaneously, monitor their activities, respond to needs, a…

Grok Voice boasts state-of-the-art performance with human-like timing, tone, and warmth in speech, offered at a significantly lower price point than…

Running coding agents locally using llama.cpp and OpenCode is highlighted as a fast, reliable, and private method for inference, addressing frustrati…