Intel has released OpenVINO 2026.1, a quarterly feature update to its AI inference optimization toolkit. Key additions include a preview OpenVINO backend for llama.cpp enabling optimized inference across Intel CPUs, GPUs, and NPUs, with validation on GGUF models like Llama-3.2, Phi-3, Qwen2.5, and Mistral-7B. The release also adds Qwen3 VL support for both CPU and GPU, GPT-OSS 120B support on CPU, and official hardware support for Wildcat Lake SoCs and the Intel Arc Pro B70 32GB GPU.
Sort: