Brief

Apple has released MLX LM Server, a new tool designed to enhance the performance of large language models on Mac hardware. It leverages the M5 chip's neural accelerators for faster prompt processing and employs continuous batching to manage multiple requests concurrently. For extremely large models, the server supports distributed inference across multiple Macs using Thunderbolt RDMA. AI

Google Fi has announced significant upgrades to its international roaming services, including faster 5G speeds and reduced prices for data usage abroad. These enhancements aim to make international travel more convenient and affordable for its users. The changes are part of Google Fi's ongoing efforts to improve its global connectivity offerings. AI

Researchers have developed OpenOpt, an open-source framework for optimizing SRAM architecture and transistor sizing. This framework utilizes equivalent circuit models to achieve significant simulation speedups while maintaining high accuracy for read/write delays and power consumption. The system integrates various optimization algorithms and has demonstrated substantial improvements in static noise margin, area, and peak power. AI

Researchers have introduced AlignFed, a new framework designed for asynchronous federated fine-tuning of large language models (LLMs) in edge environments. This approach addresses challenges like data privacy, resource heterogeneity, and non-IID data by enabling collaborative model adaptation without raw data exposure. AlignFed utilizes a multi-stage semantic alignment mechanism to mitigate model drift and aggregation fairness issues, aiming for stable and efficient LLM optimization in complex edge settings. AI

Anthropic has released two minor updates to its TypeScript SDK for Amazon Bedrock. Version v0.30.1 and v0.30.0 were pushed to GitHub, with the latter being a prerequisite for the former. These updates likely contain bug fixes or minor improvements to the SDK's functionality. AI

OpenClaw has updated its integration with Anthropic's Claude models, allowing direct API access and the reuse of Claude CLI logins. This update enables features like prompt caching and the 1 million token context window for Claude Opus 4.7. Additionally, OpenClaw now automatically handles image and PDF understanding capabilities when using Anthropic's models. AI

A new tool called 'Is It Agent Ready?' allows website owners to scan their sites for compatibility with AI agents. The tool checks for adherence to emerging standards related to discoverability, content accessibility, bot access control, protocol discovery, and commerce. It provides recommendations for improvement, such as publishing a valid robots.txt file with AI bot rules and sitemap directives. AI

This article details the migration of Airbnb's large-scale metrics pipeline from StatsD to OpenTelemetry and Prometheus. The move was driven by the need for a more robust and scalable solution to handle the increasing volume of data. The new system leverages OpenTelemetry for data collection and Prometheus with vmagent for storage and querying, improving observability and performance. AI

Relvy, a startup from the Y Combinator (YC) Winter 2024 batch, has launched its on-call runbook automation product. The platform aims to streamline incident response by providing automated runbooks. This launch targets engineering and operations teams. AI

A new project called Nanocode has been released, aiming to provide a high-performing Claude Code solution for $200. The project is built using JAX and is optimized for TPUs, suggesting a focus on efficient and powerful execution. AI

Orloj has released an open-source infrastructure-as-code platform for managing multi-agent AI systems. The tool allows developers to define agents, tools, models, memory, and other components using YAML and GitOps principles. Orloj aims to provide a declarative stack for building, operating, governing, and observing complex agentic systems, treating them like traditional software infrastructure. AI

Kita, a startup founded by Carmel and Rhea, has launched a new product designed to automate credit review for lenders in emerging markets. The system utilizes Visual Language Models (VLMs) to process diverse and often unstandardized financial documents, a task that current OCR and document AI tools struggle with. Kita's platform extracts structured financial data, detects fraud, and verifies information through cross-document checks and historical data, aiming to improve the speed and accuracy of underwriting. AI

IonRouter has launched a new inference service designed for high throughput and low cost, utilizing its proprietary IonAttention engine. This engine is capable of multiplexing multiple models on a single GPU, enabling rapid model switching and real-time traffic adaptation. The service supports various open-source models and fine-tunes, offering per-second billing and minimal cold start times, making it suitable for applications like robotics and real-time video analysis. AI

Axe is a new command-line interface tool designed to manage and execute AI agents, drawing inspiration from Unix philosophy for focused, composable functionality. It allows users to define agents with specific skills using TOML files, enabling them to be chained together or triggered by standard system tools like cron or git hooks. Axe supports multiple LLM providers, including Anthropic and OpenAI, and offers features such as persistent memory, sub-agent delegation, and structured JSON output for scripting. AI

Sentrial has launched a new platform designed to detect and alert users about failures in AI agents before they impact end-users. The service aims to provide a proactive monitoring solution for AI-driven applications. This tool focuses on identifying issues within AI agent workflows, offering a layer of reliability for businesses integrating these technologies. AI

Klaus has launched OpenClaw, an AI agent designed to function as an employee for businesses. This tool aims to simplify the integration of AI features, allowing companies to deploy agents within minutes. OpenClaw offers various use cases, including calendar management, email triage, travel booking, and meeting preparation, with tiered pricing plans and a managed rollout option for full-service deployment. AI

IonRouter has launched a new inference stack called IonAttention, designed to multiplex models on a single GPU for high throughput and low cost, compatible with NVIDIA Grace Hopper. Separately, RunAnywhere has released RCLI, an on-device voice AI for macOS that runs inference locally on Apple Silicon using their proprietary MetalRT engine, offering features like local RAG and VLM capabilities. AI

A new open-source TypeScript library has been released to help developers comply with Article 12 of the EU AI Act. This library automatically records AI inferences as tamper-evident logs, chaining entries with SHA-256 hashes and ensuring a minimum retention period. It is designed for Node.js applications using the Vercel AI SDK and aims to provide a more robust auditing solution than standard logging practices. AI

A developer has created VMPrint, a novel typesetting engine that operates without a browser, utilizing pure JavaScript for PDF generation. This engine treats document layout as a deterministic spatiotemporal simulation, where elements are autonomous actors negotiating geometry. The system is designed for high-volume report generation, collaborative editors, and print-on-demand services, offering a more efficient and reliable alternative to browser-based solutions. AI

OpenSwarm is a new command-line interface tool designed to orchestrate multiple AI agents for autonomous code-related tasks. It can integrate with various AI models, including Anthropic's Claude, OpenAI's GPT and Codex, and local open-source models. The tool aims to automate workflows such as picking up issues from platforms like Linear, running code review pipelines, and maintaining long-term memory through databases like LanceDB. AI

A new tool called Strava for Claude Code has been released, designed to help developers track their usage and costs associated with AI models like Claude. The tool provides metrics on token consumption, iteration speed, and daily usage streaks, aiming to foster a competitive environment among AI-powered developers. It emphasizes privacy by only sending aggregated usage data, not the content of prompts or code, to its local telemetry service. AI

A new open-source terminal application called Skill has been developed to facilitate the use of AI coding agents. This tool is designed to help users spin up virtual machines and GPUs, streamlining the process of deploying and managing AI development environments. The project aims to provide a next-generation development experience for those working with AI-powered coding assistants. AI

Modelence, an AI startup, has launched an open-source full-stack framework designed for both human developers and AI coding agents. The framework utilizes TypeScript for its type safety and MongoDB for flexible schema management, aiming to streamline app development by handling boilerplate tasks like authentication and database setup. An integrated app builder allows users to generate applications from prompts, with plans to introduce a DevOps agent for production monitoring and error resolution. AI

AgentMail, a new API service from Haakam, Michael, and Adi, provides dedicated email inboxes for AI agents, aiming to streamline autonomous task completion. The service addresses limitations found in existing email platforms like Gmail, offering features such as programmatic inbox creation, advanced semantic search, and usage-based pricing. Early adopters are already utilizing AgentMail for tasks like data conversion, negotiation, and training model data sourcing. AI

ShapedQL has been introduced as a new SQL engine designed to optimize multi-stage ranking and Retrieval-Augmented Generation (RAG) processes. This tool aims to streamline complex data operations within AI applications. The announcement was made via a Show HN post, indicating a focus on community feedback and developer adoption. AI

A new open-source command-line interface (CLI) and MCP server has been released to manage cloud GPU instances from Lambda. The tool, developed by Strand-AI, allows users to directly control GPU infrastructure via terminal commands or enable AI assistants like Claude to manage these resources. It offers features such as starting, stopping, and listing instances, alongside automatic notifications for instance availability across Slack, Discord, and Telegram. AI

A developer has open-sourced a comprehensive B2B SaaS starter kit built with Next.js 16 and Go 1.25. The kit includes features like enterprise-grade authentication, multi-tenancy, role-based access control, and billing integration. It also incorporates AI capabilities such as RAG pipelines with vector embeddings and an OCR service for document data extraction. AI

MCPShark is a newly released traffic inspector designed for the Model Context Protocol (MCP). This tool allows developers to observe and debug MCP traffic, including requests, responses, and tool usage, between their editor or LLM client and MCP servers. It also offers optional "Smart Scan" checks to identify potentially risky tool configurations. AI

A software engineer detailed their frustrating experience attempting to resolve a $24 Azure billing issue that prevented them from spending thousands on new services. Despite numerous attempts through official channels, including a custom-built PowerShell application, Microsoft's support system created a loop where paying the invoice required support, but support required a paid plan that couldn't be purchased due to the outstanding invoice. The engineer expressed disbelief at the company's inability to accept payment, while other users shared similar anecdotes and suggested alternative approaches like contacting sales. AI

Clad Labs has launched a new platform designed to orchestrate multiple AI coding agents, including Claude Code, Cursor, and OpenAI Codex. The tool allows developers to spin up teams of parallel agents, manage their work in isolation, and merge changes seamlessly. It also offers analytics to track coding habits and productivity alongside entertainment usage. AI

Extend, a Y Combinator-backed startup, has launched a production-ready platform designed to transform messy documents into structured data. The service utilizes specialized vision models for accurate parsing and extraction, offering features like confidence scoring, multiple processing modes, and an optimization agent called Composer Agent to refine schemas automatically. Extend aims to streamline document workflows for AI teams, enabling faster development and deployment of data processing pipelines. AI

AI startup Webhound has launched a research agent designed to automate the creation of web-scraped datasets based on natural language prompts. The agent, initially built on Claude 4 Sonnet, was re-engineered using Gemini 2.5 Flash and a multi-agent system to significantly reduce costs and improve reliability. This new architecture includes specialized agents for planning, searching, critiquing, and validating data, along with a text-based browser for efficient extraction. AI

A developer has created an open-source AI-powered web service that integrates FastAPI for APIs, Pydantic-AI for agent construction, and Model Context Protocol (MCP) servers for tools. The service allows users to query information from sources like Hacker News and web search, presenting ranked trend cards with summaries. It supports various local LLM configurations and is containerized with Docker for production deployment. AI

Recall.ai has launched a new Desktop Recording SDK designed to simplify the integration of meeting recording capabilities into other applications. This SDK addresses the complexities of capturing high-quality audio and video, including speaker identification and clean video compositing, without requiring a bot to be present in the meeting. The company aims to provide developers with a robust infrastructure solution, drawing on their experience powering recording features for over 2000 companies and overcoming significant technical challenges in reliability and efficiency. AI

Smooth has launched a new serverless browser agent API designed for reliability, speed, and cost-efficiency, claiming to be 7x cheaper and 5x faster than existing solutions. The API aims to simplify web automation tasks for developers by handling complexities like instant browser spin-up and CAPTCHA solving. Separately, ContextFort has introduced a tool to provide visibility and control over AI coding agents like Cursor and Claude Code, addressing security concerns about agents accessing sensitive files and credentials on developer machines. AI

April, a new voice-controlled AI assistant, has launched on the App Store to manage emails and calendars. The application allows users to dictate replies, summarize messages, and reschedule meetings hands-free. It utilizes Deepgram for speech-to-text and Eleven Labs for text-to-speech, with custom servers for Google integration. The developers are focusing on low latency and natural interaction, while also considering user feedback on safety features like a 'safe mode' for non-destructive operations. AI

Skope, a new billing system, has launched to support outcome-based pricing for software products, particularly targeting the burgeoning AI market. The platform allows companies to charge customers only when their software delivers a specific result, aligning incentives and reducing buyer risk. Skope aims to simplify the implementation of this pay-per-performance model, which was previously challenging to manage at scale. AI

Cyberdesk, a startup founded by Mahmoud and Alan, has launched a new tool designed to automate repetitive tasks within legacy Windows desktop applications. Their approach uses a deterministic computer use agent that learns workflows from natural language instructions, offering a more reliable alternative to traditional Robotic Process Automation (RPA) scripts. The agent can self-correct based on screen state and only resorts to expensive AI models when unexpected anomalies occur, making it both robust and cost-effective for industries like healthcare and accounting. AI

The mcp-use framework has been released, enabling developers to build applications that can connect to various large language models like ChatGPT and Claude. This framework allows for the creation of MCP Servers and MCP Apps, with SDKs available in TypeScript and Python. It also includes an MCP Inspector for testing and debugging, and a cloud deployment option for production environments. AI

A new infrastructure project called PHP-ORT aims to bring machine learning inference capabilities directly to PHP, the server-side language used by a significant portion of the web. This development seeks to empower millions of PHP developers to integrate AI features into their applications without relying on external services or switching programming languages. PHP-ORT provides a core Tensor API, a high-performance math library, and integrates with ONNX for direct inference, promising significant speedups. AI

ZeroEntropy has developed specialized AI models, including rerankers and embeddings, designed for production systems that prioritize speed and accuracy over generalist models. Their offerings, such as zembed-1 and zerank-2, aim to provide lower latency and higher accuracy for applications like Retrieval Augmented Generation (RAG). These models are available for integration into existing stacks and can be deployed on cloud platforms like AWS and Azure, with a focus on security and compliance standards. AI

BrowserOS has launched as an open-source browser designed for the AI era, integrating AI agents that can automate web tasks through natural language commands. It prioritizes user privacy and offers extensive customization by supporting over 11 AI providers, including popular options like Anthropic Claude, Google Gemini, and OpenAI, as well as local models. The browser is built on a Chromium fork, ensuring compatibility with existing Chrome extensions and offering a user-friendly experience for both general users and developers. AI

Glowstick is a new Rust crate designed to enhance tensor manipulation by integrating shape checking directly into the type system. This approach aims to make tensor operations safer and more intuitive, particularly for developers working with machine learning frameworks. The project, currently in its pre-1.0 phase, offers features like dynamic dimension support and improved error messages, with plans to align with ONNX operations. AI

Tinfoil, a startup founded by researchers from MIT and Cloudflare, has launched a new service designed to provide verifiable privacy for AI workloads hosted in the cloud. The platform utilizes secure enclave technology, particularly NVIDIA's confidential computing capabilities on GPUs, to ensure that neither Tinfoil nor the cloud provider can access sensitive data processed by AI models. This approach aims to enhance AI privacy by replacing trust with provable security, enabling more complex AI applications that require private data. AI

ParaQuery, a new startup, has launched a GPU-accelerated Spark and SQL data processing solution. The platform aims to offer cost and performance benefits over existing solutions like Google BigQuery. ParaQuery leverages NVIDIA's RAPIDS technology to enhance traditional data processing tasks, which the founder notes are often mistakenly believed to be limited to AI and graphics. AI

Exa has launched Websets, a new search engine that uses embeddings and agentic workflows to provide precise results from the web, presented in a database-like table format. The service aims to combat the decline in search quality by performing extensive embedding searches and then using LLMs to verify each result against complex queries. While the process can take significant time, Exa believes the accuracy and detailed verification are worth the wait, offering an alternative to traditional keyword-based search. AI

Raven is a new ecosystem of OCaml libraries designed for numerical computing, machine learning, and data science. It aims to provide type-safe alternatives to popular Python libraries such as NumPy, JAX, and PyTorch. The project includes modules for n-dimensional arrays, automatic differentiation, tokenization, neural networks, dataframes, and plotting, with the goal of building a robust scientific computing environment. AI

Morphik has launched an open-source Retrieval-Augmented Generation (RAG) system designed for developers to integrate complex context into AI applications. The system aims to simplify the process by offering a unified solution for storing, representing, and searching unstructured and multimodal data, addressing the limitations of traditional RAG pipelines that struggle with visually rich documents. Morphik provides features like multimodal search, fast metadata extraction, and integrations with tools such as Google Suite and Slack, with a free tier available for users. AI

A new open-source project offers sandboxed Chrome browsers that can be run as Docker containers or on Unikraft unikernels. This setup is designed for browser automation, web agents, and testing AI agents that interact with the web. The unikernel implementation provides features like automated standby mode with state snapshotting and extremely fast cold restarts, enabling low-latency event handling. AI

AI startup mrge has launched a new platform designed to streamline code reviews for development teams. The tool connects to GitHub repositories and uses AI to analyze code changes within a secure, ephemeral sandbox environment. It aims to assist human reviewers by identifying potential bugs and providing context, inspired by productivity tools like Linear and Superhuman. AI