Replicate is building tools so all software engineers can use AI as if it were normal software. You should be able to import an image generator the same way you import an npm package. You should be able to customize a model as easily as you can fork something on GitHub.
Find out more about Text To Image here.
from huggingface_hub import InferenceClient
client = InferenceClient(
provider="replicate",
api_key="hf_xxxxxxxxxxxxxxxxxxxxxxxx",
)
# output is a PIL.Image object
image = client.text_to_image(
"Astronaut riding a horse",
model="black-forest-labs/FLUX.1-dev",
)Find out more about Text To Video here.
from huggingface_hub import InferenceClient
client = InferenceClient(
provider="replicate",
api_key="hf_xxxxxxxxxxxxxxxxxxxxxxxx",
)
video = client.text_to_video(
"A young man walking on the street",
model="Wan-AI/Wan2.1-T2V-14B",
)