Ming-UniVision is a large multimodal language model that first integrates continuous visual representations into the next-token prediction framework, unifying vision and language under a single autoregressive paradigm without discrete quantization or modality-specific heads. This model supports joint image understanding and generation, converges faster in vision-language training, and also supports multi-round context visual tasks.

MultimodalTransformersMultiple Languages

GLM-4.5-Air is a foundation model for agents launched by Zhipu AI. It adopts a compact design, unifying reasoning, coding, and agent capabilities to meet the complex requirements of agent applications. The model has a moderate number of parameters and superior efficiency. Multiple versions have been open-sourced and can be used for commercial purposes and secondary development.

Natural Language ProcessingTransformersMultiple Languages

Florence-2 is an advanced visual foundation model developed by Microsoft, employing a prompt-based architecture to unify various vision and vision-language tasks

MultimodalSafetensors

RADIO is a vision foundation model developed by NVIDIA Research, capable of unifying visual information across different domains for various vision tasks.

Computer VisionTransformers

ChatTime is an innovative multimodal time series foundation model that treats time series as a foreign language, unifying the processing of bimodal input/output for both time series and text.

MultimodalTransformers

A general-purpose image segmentation model based on Swin Transformer, unifying instance/semantic/panoptic segmentation tasks

Computer VisionTransformers

A large-scale Mask2Former model based on the Swin backbone network, designed for general image segmentation tasks, unifying instance segmentation, semantic segmentation, and panoptic segmentation.

Computer VisionTransformers

Large-scale MaskFormer model based on Swin backbone network, unifying instance/semantic/panoptic segmentation tasks

Computer VisionTransformers

MaskFormer semantic segmentation model trained on the ADE20k dataset, using a Swin backbone network to unify instance/semantic/panoptic segmentation tasks

Computer VisionTransformers

A panoptic segmentation model based on the Swin backbone network, trained on the COCO dataset, unifying instance/semantic/panoptic segmentation tasks

Computer VisionTransformers