https://onurbaydas.com/en/tags/inference/Recent content in Inference on Onur BaydaşHugoen-usSat, 17 Oct 2026 12:15:22 +0300https://onurbaydas.com/en/news/2026-10-17-gemini-4-argon-frontier-intelligence/Sat, 17 Oct 2026 12:15:22 +0300https://onurbaydas.com/en/news/2026-10-17-gemini-4-argon-frontier-intelligence/Google DeepMind announced a new generation AI model named Gemini 4 Argon. This model aims to push the capabilities defined by the company as "frontier intelligence" a step further.https://onurbaydas.com/en/news/2026-10-17-nvidia-rtx-spark-ai-notebook/Sat, 17 Oct 2026 12:15:22 +0300https://onurbaydas.com/en/news/2026-10-17-nvidia-rtx-spark-ai-notebook/An Nvidia RTX Spark-powered notebook delivers significant compute power for local AI workloads. This device brings high-performance AI capabilities in a mobile form factor.https://onurbaydas.com/en/news/2026-10-17-open-weight-models-vercel-usage/Sat, 17 Oct 2026 12:15:22 +0300https://onurbaydas.com/en/news/2026-10-17-open-weight-models-vercel-usage/Open-weight AI models account for 71.5% of tokens processed through the Vercel gateway. These models represent nearly a quarter of estimated spending, indicating growing adoption.https://onurbaydas.com/en/news/2026-10-16-intel-gpu-d-series-crescent-island/Fri, 16 Oct 2026 00:17:33 +0300https://onurbaydas.com/en/news/2026-10-16-intel-gpu-d-series-crescent-island/Intel has unveiled the official performance specifications and branding for its "Crescent Island" codenamed data center AI inference cards. These new products will be known as the Intel GPU D-Series.https://onurbaydas.com/en/news/2026-10-16-lfm25-24b-a2b-on-device-agents/Fri, 16 Oct 2026 00:17:33 +0300https://onurbaydas.com/en/news/2026-10-16-lfm25-24b-a2b-on-device-agents/Hugging Face announced the LFM2.5-24B-A2B model. This model enables AI agents to run directly on-device, offering fast and private operations.