# High-precision Vision

AIMv2 is a series of vision models pretrained with multimodal autoregressive objectives, demonstrating outstanding performance across multiple multimodal understanding benchmarks.

High-performance vision embedding model, sharing the same embedding space with nomic-embed-text-v1, supporting multimodal applications

BEiT base model fine-tuned on an unknown dataset, specific use cases and performance details are currently unavailable