How to run Ming-Lite-Omni locally

inclusionAI (Ant Group) · Omni understanding + image & speech gen · 19B params · MIT (commercial OK)

An MoE omni model (19B total, only 2.8B activated) that ingests image, text, audio and video and uniquely generates both images (with editing and style transfer) and speech, all under a permissive MIT license.

What it costs to run — $0 markup

On your own machine — $0. Runs free if you have about 41 GB of VRAM (~41GB in BF16 (19B total, 2.8B active MoE) needs an A6000/A100-class 48GB card at full precision; quantization can fit a 24-32GB card.), via Transformers (BailingMM).
Rent a GPU — from $0.49/hr. Fits RTX A6000 48GB at the direct vendor price ($0 markup) — pay only for the minutes you generate.
Download the weights — free. Open weights at inclusionAI/Ming-Lite-Omni.

Note: generative-media models are billed per image / per second / per minute on hosted services — not per token. Running locally or on your own rented GPU is usually far cheaper and keeps your data on your machine.

Key facts

Does	Visual understanding, Audio understanding, Video understanding, Text → image, Image editing, Text → speech, Speech recognition
VRAM to run	~41 GB (~41GB in BF16 (19B total, 2.8B active MoE) needs an A6000/A100-class 48GB card at full precision; quantization can fit a 24-32GB card.)
Download	~41 GB
Parameters	19B
License	MIT (commercial use OK)
Run with	Transformers (BailingMM)

Get Ming-Lite-Omni on Hugging Face →

More multimodal / omni models

Qwen2.5-Omni-7B — Omni understanding + speech generation, ~24 GB VRAM
Janus-Pro-7B — Unified understanding + image gen, ~16 GB VRAM
OmniGen2 — Any to image (gen + edit + in context), ~17 GB VRAM
MiniCPM-o 2.6 — Omni understanding + speech generation, ~9.0 GB VRAM
BAGEL-7B-MoT — Unified understanding + image gen + edit, ~24 GB VRAM
Emu3.5 — Any to any world model (gen + edit), ~48 GB VRAM

Browse: all media models · chat / LLM models