Full-text search across 4717 articles. Combine with topic and date filters; results sorted by relevance.
Alibaba released Qwen 3.8 Max, a 2.4 trillion parameter open-weight AI model with 95 billion active parameters, set to become the largest open-weight model once weights are open-sourced about a week after announcement. T…
Alibaba made its Qwen3.8-Max model generally available on August 3, 2026, a 2.4-trillion-parameter Mixture-of-Experts model with a 1M-token context window and native text, image, and video input. Priced at $2 per million…
Exa CEO Will Bryk said on August 3rd that Exa's index now serves 80 billion pages and tracks 1.4 trillion URLs, nearly triple the 500 billion URLs reported in May, with expectations to reach "Google-scale" by early 2027.…
Building a plugin for Until, a plan-first workflow for coding agents, revealed that agent-first products have three interfaces, and the agent interface is where most failures occur. The team found that instructions alone…
Keystroke, a Y Combinator W24 company that pivoted from Buster, has open-sourced its internal agents and automations platform under the ELv2 license. The platform allows developers to build agents and durable workflows a…
A developer argues that integrating the Language Server Protocol (LSP) into AI coding agents is essential for reliable code understanding, moving beyond brittle prompt-based context selection. The approach enables determ…
Superblocks launched version 3.0 of its AI app-building platform on Monday, introducing a deployment model that runs the entire platform, including AI inference, inside a customer's Amazon Web Services (AWS) environment.…
Baseten, an AI infrastructure company that recently raised a $1.5B Series F led by Altimeter Capital, Conviction Partners, and Spark Capital, is hiring an AI Inference Engineer in San Francisco with a salary range of $16…
Markov, a computer-use data startup founded by Dev Mandal and Harish Ashok, released CAD Environments on August 3rd, a dataset of 51 human-performed CAD workflows totaling 99.03 hours and 152 GB for training and evaluati…
A developer at AWS describes a new approach to learning web development in 2026, emphasizing the use of AI agents to accelerate the process. The method involves building a personal project, learning only the core primiti…
A freelance developer's experiment found that building custom charts with D3 primitives and React matched a client's design system exactly, while a Recharts-based version required custom SVG code for the final 20% and hi…
In a blog post, technology writer and designer Matt Webb proposed a voice-first podcast app that lets listeners talk back to the show to capture notes, citing a need for a minimal Granola-like interface to timestamp mome…
MinerU, a high-accuracy document parsing engine for LLM, RAG, and Agent workflows, released version 3.4 on 2026/06/18, upgrading the pipeline backend's OCR model to PP-OCRv6, which improves OCR accuracy by about 11% on O…
A developer building claude-real-video, an MIT-licensed local pipeline for feeding videos to vision LLMs, found that frame selection is the critical factor in video understanding, as LLMs can afford about 150 images per …
GitHub has released gh-stack, a GitHub CLI extension that automates stacked pull requests by managing branches, rebasing, and PR base branches. The tool, installable via `gh extension install github/gh-stack`, requires G…
JetBrains, the software development company, reported that its AI development expenses increased roughly 10x over the past six months, prompting the company to implement new cost-control measures. The spike, starting in …
Cecuro launched Ozone, an AI security engineer that reviews every GitHub pull request in an isolated sandbox, with the first $100 of reviews free. The tool, built on the same multi-agent engine as Cecuro's smart contract…
A developer's analysis argues that the real bottlenecks in AI coding assistants are 'cognitive debt'—the mental overhead of verifying AI-generated code—and technical limitations in Language Server Protocol (LSP) integrat…
Congress is pressing DoorDash to disclose its use of Chinese AI models, including Moonshot AI's Kimi, after co-founder Andy Fang said the company routes lower-level engineering work to Kimi K2.6. House Select Committee o…
Newsroom, a self-hosted, AI-ranked meta-aggregator for Hacker News, Substack, podcasts, Bluesky, Reddit, and other feeds, was released on GitHub. The Node.js-based tool combines keyword and optional AI ranking to create …