OpenRouter has tracked the deepseek/deepseek-v4-flash-0731 endpoint at streamlake/fp8 since 2026-08-13, recording $0.032 in total spend across 66,386 queries. The B3IT border-input detector measured a…
Inception released Mercury 2.5 Preview on August 31, 2026, claiming 1,107 tokens per second on standard GPUs, a 10-point intelligence jump over Mercury 2, and quality comparable to GPT-5.6 Luna at low…
A developer has built GoleTalk, a Somali-first social platform designed to serve Somali-speaking communities with language-aware features and integrated AI capabilities. The platform uses Flutter, Lar…
AI startup TrustedRouter has raised $1.25 million, according to an Axios report. The company offers a privacy-focused router for AI services, allowing users to avoid sharing data with third-party prov…
Data Convo, an AI analytics platform, announced it can be deployed on AWS, Azure, or private servers in under two minutes, allowing enterprises to run its agentic workspace on their own infrastructure…
Zhipu AI's open-weight GLM 5.3-Flash, released under an MIT license, processed roughly 23 trillion tokens in its first week on OpenRouter, becoming the platform's most-used model. With 320B parameters…
A developer tested five free LLM APIs—Groq, Google AI Studio, OpenRouter, Cloudflare Workers AI, and Token Harbor—using a single Python script and two tasks. The results, published on August 28–29, 20…
A production audit of 10,186 assistant turns across OpenAI, Google Gemini, and OpenRouter found that dynamic tool activation causes prompt cache misses on only 2.4% to 3.4% of turns, while avoiding 71…
Mrigashira AI, a desktop AI engine for running large language models locally, is now available on Linux distributions. The tool dynamically adjusts context windows based on available RAM/VRAM to suppo…
Preseason, an open-source benchmark launched on Hacker News, measures which developer tools large language models recommend when asked to build real web apps, tracking recommendations across categorie…
A new census of open-weight model hosting reveals that the same model name can arrive at different quantizations, context lengths, and output caps depending on the provider, with some hosts serving GL…
Aniket Wathore has published Bolnee-Chat, an open-source retrieval-augmented chatbot for business websites, on GitHub. The self-hosted tool, which had 43 commits and one star as of August 30, lets ope…
Chinese AI company Z.ai confirmed on 26 August that the anonymous AI model Ox Alpha, which appeared on OpenRouter on 20 August, was a stealth preview of its GLM-5.3-Flash model. The model, which attra…
Pangolin, the open-source zero-trust remote access platform, released version 1.22, adding an AI Gateway that acts as an identity-aware proxy between AI clients and model providers, supporting OpenAI,…
Bolnee-Chat, a self-hosted RAG chatbot platform, is now available for free on GitHub, enabling businesses to integrate AI chatbots into their websites with no vendor lock-in or per-message billing. Th…
Tencent released Hy4 preview on August 28, a 770B parameter open-weight model under Apache 2.0, with only 49B active parameters per token and a 1 million token context window, targeting software engin…
OpenRouter, an AI model aggregator, lists popular Chinese models such as DeepSeek, Z.ai, and MiniMax as headquartered in Singapore on its provider directory, despite their Chinese origins. The directo…
Tencent released Hy4 Preview, a large language model with a significant size increase from its predecessor Hy3, which had 295B total parameters, 21B active parameters, 256,000 context length, and 598G…
OpenRouter retired monitoring of the deepseek/deepseek-v4-flash-0731 endpoint at akashml/fp8 on 2026-09-11 after it stopped yielding usable samples, leaving its B3IT tracking status retired and stalle…
Monitoring of the z-ai/glm-5.2 endpoint at gmicloud/fp8 was retired as of 2026-09-04 because a single query costs more than the per-query guard allows, according to the tracking dashboard. The endpoin…