Local Ai

v0.31.2-rc2: llm: allow iGPU mmproj offload with fit padding (#16996)

This release candidate introduces support for offloading the image GPU projection (mmproj) to an integrated GPU when using fit padding in Ollama's LLM processing, addressing technical improvements for

DGX agentgithub
local-aiollama-releases

This release candidate introduces support for offloading the image GPU projection (mmproj) to an integrated GPU when using fit padding in Ollama's LLM processing, addressing technical improvements for GPU memory management and multimodal model inference on systems with integrated graphics.

Source: Ollama Releases | 2026-07-07

Loading related sources…