The AI-revolution is amongst us. This means an explosion in content created by tools like Stable Diffusion, MidJourney, Dall-E, and many many more. Something I've been asking myself is where is everyone storing these? Today it's just pictures, but soon it will be videos, texts, even powerpoint presentations.

According to OpenAI:

More than 3 million people are already using DALL·E to extend their creativity and speed up their workflows, generating over 4 million images a day.

I've been building Mixpeek as an "intelligence layer" on top of traditional cold storage, serving up headless version control, indexing, searching, and analytics for developers that want to enhance their application's search experience.

This will be a quick tutorial in which we will use HuggingFace's stableDiffusion library to generate an image, store it in Mixpeek's indexer, then quickly retrieve it.

Generate the Image

from diffusers import StableDiffusionPipeline, DPMSolverMultistepScheduler

model_id = "stabilityai/stable-diffusion-2-1"

pipe = StableDiffusionPipeline.from_pretrained(model_id, torch_dtype=torch.float16)

pipe.scheduler = DPMSolverMultistepScheduler.from_config(pipe.scheduler.config)
pipe = pipe.to("cuda")

image = pipe("a photo of an astronaut riding a horse on mars").images[0]

image.save("astronaut_rides_horse.png")

Index the Image

from mixpeek import Mixpeek

mix = Mixpeek("YOUR_API_KEY")

# store the stable diffusion output in your own mixpeek cloud store
mix.index(
  "astronaut_rides_horse.png",
  description="a photo of an astronaut riding a horse on mars",
  tags="stable-diffusion, space, animal, scifi",
  save=True
)

Search your Assets

mix.search("space horse")

[
  {
    "file_id": "63a86c1bf1fc46279ef7614d",
    "filename": "astronaut_rides_horse.png",
    "importance": "95%"
  },
  {
    "file_id": "63a86c21f1fc46279ef7614e",
    "filename": "cowboy_on_mars.png",
    "importance": "72%"
  }  
]

On top of search we have the following capabilities across the tooling and analytics API set to make your life easier:

  • Analytics - See which searches are most popular, which yield no results, and which lead to a conversion (clickthrough)
  • Deduplication - Find similar appearing images to reduce redundancies
  • Alerts - Become notified when similar assets are added to your Mixpeek directory
  • Many, many more.
Mixpeek is a multimodal indexing and search service. We offer managed APIs that are faster, cheaper, and more relevant than the competition all while providing impressive tooling on top.
About the author
Ethan Steininger

Ethan Steininger

Former GTM Lead of MongoDB's NLP platform, Atlas Search. Occasionally off the grid in his self-converted camper van.

Multimodal Makers | Mixpeek

Multimodal Pipelines for AI

Great! You’ve successfully signed up.

Welcome back! You've successfully signed in.

You've successfully subscribed to Multimodal Makers | Mixpeek.

Success! Check your email for magic link to sign-in.

Success! Your billing info has been updated.

Your billing was not updated.