Meta Prometheus on BriefTechNews.comhttps://brief-tech-news.com/tags/meta-prometheus/Recent content in Meta Prometheus on BriefTechNews.comHugoenCopyright &copy;2021 <br> Designed &amp; Developed by [Gethugothemes](https://gethugothemes.com/)Tue, 20 Oct 2026 13:46:52 +0000MiniMax Cuts Agent Costs 70% as Mercury 3 Hits 1,068 Tokens/Secondhttps://brief-tech-news.com/blogs/2026-10-20-ai/Tue, 20 Oct 2026 13:46:52 +0000https://brief-tech-news.com/blogs/2026-10-20-ai/MiniMax launched M3.1 Flash with nearly 3× faster online inference, 70% lower costs and stronger coding, computer-use and security scores. Inception Labs’ Mercury 3 generated 1,068 tokens per second with a 1-million-token context, while Cloudflare previewed Clairic, a token-free model for extracting typed values. Meta detailed a training supercluster heading toward 1 million GPUs, and AWS reported 3× better time-to-first-token from push-based KV-cache transfers in vLLM.