We used HFlow to evaluate the latest open weights VLMs for processing egocentric data

We used HFlow to evaluate the latest open weights VLMs for processing egocentric data. This was based on Build AI's Egocentric-10k evaluation , which used Gemini 2.5 Flash to measure hand visibility and active manipulation. We kept the same prompts and the same dataset, only varying the model. How much each model agreed with the original results: Gemini 2.5 Flash: 91.65% (baseline) GLM 5.3 Flash: 91.00% Gemma 4 26B-A4B: 90.87% Qwen 3.8 27B: 90.79% Inkling Small: 85.21% Gemma was the standout. Its results were on par with Gemini while being 19x cheaper. Both Gemma and Qwen models are practical to self-host, enabling private processing without data egress. This suggests modern open weights VLMs are becoming good enough for large-scale egocentric data processing. The main differentiators are increasingly cost, throughput, output reliability, and ease of self-hosting. If you're optimizing multimodal processing for egocentric data, you can run this evaluation yourself with any prompt and model using Hflow. github.com/Hebbian-Robotics/hflow git clone github.com/Hebbian-Robotics/hflow.git cd hflow/examples/build_ai_evaluation