Vision-language generation

Family-owned implementations of the vision-language generation task interface.

See the related Hugging Face task page for the ecosystem-level task definition.

Model families

Model family	Declared recipes	Exact checkpoint examples	Runtime CLI
`deepseek_ocr`	4	`deepseek-ai/DeepSeek-OCR-2`	`trtmc run`
`internvl`	5	`OpenGVLab/InternVL3-2B-hf` `OpenGVLab/InternVL3-8B-hf`	`trtmc run`
`lance`	1	`bytedance-research/Lance`	`trtmc run`
`locateanything`	2	`nvidia/LocateAnything-3B`	`trtmc run`
`nemotron_omni`	1	`nvidia/Nemotron-3-Nano-Omni-30B-A3B-Reasoning-NVFP4`	`trtmc run`
`phi4_multimodal`	1	`microsoft/Phi-4-multimodal-instruct`	`trtmc run`
`qwen_vl`	7	`Qwen/Qwen2.5-VL-3B-Instruct` `Qwen/Qwen3-VL-2B-Instruct` `Qwen/Qwen3-VL-8B-Instruct` `nvidia/Qwen2.5-VL-7B-Instruct-FP8` `nvidia/Qwen2.5-VL-7B-Instruct-NVFP4`	`trtmc run`
`qwen3_5`	2	`Qwen/Qwen3.5-27B` `Qwen/Qwen3.5-2B`	`trtmc run`
`qwen3_5_moe`	1	`Qwen/Qwen3.5-35B-A3B-GPTQ-Int4`	`trtmc run`