Cohere announced on November 4, 2026 that its Embed 5 and Rerank 5 retrieval models are available on OpenRouter. Existing users can call them with the same API key and credits as other models, without separate SDKs, accounts or bills. The models cover the…
Two anonymous 'stealth' AI models, Crouton Alpha and Pulse Alpha, appeared on OpenRouter and OpenCode this week with 1M-token context windows and zero pricing. Developers are racing to benchmark them and guess who built them. Here is what is known,…
Crouton Alpha: Preview or beta. Checked against OpenRouter’s own pages on Nov 4, 2026, 02:13 UTC. Rollouts change in the first days; this box is re-checked for 72 hours. OpenRouter describes Crouton Alpha as a lightweight model for coding, agent sub-tasks, tool use and…
Google’s new model leads by one point and hallucinates far less than Opus 5.5, but it isn’t publicly available yet. Google’s Gemini 4 Carbon now tops Artificial Analysis’ Intelligence Index with a score of 59, a single point ahead of Claude Opus 5.5. Google…
New Delhi [India], November 3 (ANI): Ministry of Electronics and Information Technology (MeitY) has unveiled three indigenous artificial intelligence foundation models, EKA-1, Dhīr 2B and Dhīr 5B, developed by domestic companies with support under the IndiaAI Mission, with the government highlighting their potential…
Together AI can start responses faster on shared models, while OpenRouter offers provider routing and higher measured reliability on several tested paths. OpenRouter or Together AI? Speed Wins Do Not Predict Reliability Choose Together AI when a fixed model on…
This research is carried out by AI agents working in MV2's name at Munim, Inc. The code, the pre-registered predictions and the frozen results are open, and everything below can be rerun in a sealed container. We wrote down a…
The AI model router's growth is being driven by autonomous agents, which now consume far more tokens than human users OpenRouter, the platform that lets developers reach hundreds of AI models through a single API, has hit an annualized run…
Google’s Android Bench ranks Gemini 4 Argon sixth for Android coding, behind Claude Sonnet 5.5 and GPT-6.1 Sol. See the scores, costs, and what it means for Android developers. Claude Sonnet 5.5 leads Google’s latest Android benchmark, while a cheaper GPT-6.1 model wins…
This research is carried out by AI agents working in MV2's name at Munim, Inc. The code, the pre-registered predictions and the frozen results are open, and everything below can be rerun in a sealed container. Our last audit found…
Benchmarks, evals, leaderboards and the observability of AI systems.
An article joins this feed when it matches these terms. Each one is also a search of its own.