Developer ClaudioDrews released Memory OS, an open-source six-layer memory stack for the Hermes Agent, under an MIT license. The system adds a vector database, structured facts, and an auto-curated kn…
SigNoz has released a guide for setting up OpenRouter observability using OpenTelemetry, enabling developers to track LLM API requests, latency, and usage across multiple models in unified dashboards.…
A developer built Keph, an uncensored AI chatbot powered by Nous Research's Hermes 4 70B model that delivers unfiltered answers on sensitive topics without refusals or moral lectures. The chatbot, acc…
A developer tested six methods to use Claude Code with free AI models, finding that only one worked reliably. The engineer encountered failures with OpenRouter's free tier due to rate limits, proxy tr…
An analysis of 60+ AI inference providers using 398 data points reveals that cache hit rates, not headline input costs, are the primary driver of real-world pricing for agentic workflows. Chinese labs…
A photographer has built AgentFlow, a declarative DSL that lets users define multi-agent AI workflows in simple `.aflow` files and compile them into MCP tools for Claude Code without writing any integ…
CompletionKit has launched an open-source tool for testing AI prompts against real inputs, scoring outputs with custom metrics, and identifying regressions before deployment. The platform supports mul…
HoneyChat, a Telegram-native AI companion with approximately 300 daily active users across 17 languages, has implemented a three-step fallback chain that recovers roughly 70% of false-positive LLM ref…
Boogy launched a production infrastructure platform that lets developers deploy full backends — including REST, RPC, MCP, sandboxed databases, and authentication — in seconds from a single prompt. The…
A developer built Council, a system that pits three different AI models against each other to debate judgment calls, then uses Hermes as a judge to deliver a single verdict with a confidence score and…
Bittensor integrated a confidential routing layer with OpenRouter through Subnet 28, processing between 100 billion and 120 billion tokens daily. The system uses Trusted Execution Environments to keep…
A new benchmark evaluating DocumentAI models on bounding box accuracy shows GPT-5.5 and Gemini 3.5 leading with 67.7% and 67.5% scores respectively, while Qwen, Kimi, and Mistral trail significantly. …
A developer fine-tuned the Qwen2.5-0.5B model on 700 real incident post-mortem examples, achieving over 60% rubric compliance for structured SRE summaries—outperforming zero-shot baselines from larger…
Meltdown, a lightweight Python-based LLM client built with Tkinter, now supports OpenRouter alongside local servers, ChatGPT, Gemini, Claude, and Kimi. The tool offers custom widgets, a scratch-built …
A maxillofacial surgeon and self-taught developer in Ouagadougou, Burkina Faso shipped Readium, a production AI reading app that lets users discuss books with Claude in any language. The developer enc…
A previously unknown AI model called Hy3 preview, developed by Chinese tech giant Tencent, has surged to the top of OpenRouter's AI Model Rankings, surpassing established models like Claude by more th…
A developer abandoned attempts to make an AI website builder generate proper desktop layouts through prompt engineering after two weeks of failed iterations. The engineer found that large language mod…
AI coding startup Cognition raised $1 billion at a $25 billion valuation, while OpenRouter doubled its valuation to $1.3 billion in one year. Microsoft shipped Copilot Cowork, an autonomous AI agent b…
A developer built ffmpeg-ai, a free Python CLI that converts a single text prompt into a finished, upload-ready 1080x1920 MP4 video for YouTube Shorts, Reels, or TikTok. The tool automates the entire …
A team running an AI companion bot measured prompt caching performance across multiple providers in production and found hit rates ranging from 0% to 91% for the same 5,000-token system prefix. Cydoni…