# Image-text generation

An INT8 quantized version based on google/gemma-3-12b-it, supporting visual text input and text output, suitable for efficient inference deployment

Jedi-3B-1080p is a 3B-parameter model developed by xlangai, quantized using llama.cpp, suitable for image-text generation tasks.

medgemma-4b-it is a multimodal model focused on the medical field, capable of processing image and text inputs, and suitable for multiple medical scenarios such as radiology and clinical reasoning.

InternVL3 is an advanced multimodal large language model series with powerful multimodal perception and reasoning capabilities, supporting image, video, and text inputs.

InternVL3-2B is a multimodal large language model implemented based on the Hugging Face Transformers library. It performs excellently in multimodal tasks such as image, video, and text processing, supporting multiple input methods and efficient batch inference.

Kimi-VL-A3B-Thinking-8bit is a multimodal vision-language model converted based on the MLX format, supporting image-text to text generation tasks.

Gemma 3 27B IT QAT BF16 is a version of the Gemma series of models released by Google. It has undergone quantization-aware training (QAT) and is converted to the BF16 format, suitable for the MLX framework.

Gemma 3 is a lightweight multimodal open model from Google, supporting text and image inputs with text output, featuring a 128K large context window and multilingual capabilities.

Gemma is a lightweight, advanced open model series from Google, built using the same research technology as Gemini. Gemma 3 is a multimodal model capable of processing both text and image inputs to generate text outputs.

This model is converted from ds4sd/SmolDocling-256M-preview to the MLX format, supporting image-text-to-text tasks.

A 72B-parameter multimodal foundation model released by ByteDance Research, specializing in image-text-to-text tasks

Aya Vision 8B is an open-weight 8-billion-parameter multilingual vision-language model supporting visual and language tasks in 23 languages.

Gemma is a lightweight open-source multimodal model series launched by Google, built on the same technology as Gemini, supporting text and image inputs and generating text outputs.

FP8 dynamically quantized model based on Aria-sequential_mlp, suitable for image-text-to-text tasks, requiring approximately 30GB VRAM.

A vision-language model based on Microsoft Florence-2-large, excelling in image understanding and text generation tasks

IDEFICS is an open-source multimodal model capable of processing both image and text inputs to generate text outputs, serving as an open-source reproduction of Deepmind's Flamingo model.

BLIP-2 is a vision-language pre-trained model that achieves language-image pre-training guidance by freezing the image encoder and large language model.