Moondream

# pip install moondream
import moondream as md
from PIL import Image

# Moondream 3 on Photon — local GPU inference
model = md.photon("moondream3-preview")

image = Image.open("photo.jpg")

# Caption the image
caption = model.caption(image)["caption"]
print("Caption:", caption)

# Ask a question
answer = model.query(image, "What's happening in this image?")["answer"]
print("Answer:", answer)

# Moondream 3 also detects, points, and segments
objects = model.detect(image, "person")["objects"]
for obj in objects:
    print("Person:", obj)