About the challenge
Create the next generation of AI media apps.
The Backblaze Generative AI Media Hackathon challenges developers to build apps for AI-generated video, images, audio, and multimodal media workflows using Backblaze B2 Cloud Storage and Genblaze.
Backblaze B2 is S3-compatible cloud object storage that helps generative media apps store and serve large volumes of generated assets, metadata, provenance records, thumbnails, logs, and application files as part of the media pipeline. Start building for free with 10GB of included B2 storage, then scale as your media workloads grow.
Genblaze is an open source Python SDK from Backblaze for orchestrating generative media workflows across generative media providers such as GMI Cloud, OpenAI, Google, Decart, Runway, Luma, ElevenLabs, Stability Audio, NVIDIA NIM, and others.
GMI Cloud gives developers easy access to open source generative AI models for image, video, audio, chat, reasoning, and multimodal workflows, making it easier to prototype and build AI media applications.
Build something useful, creative, and production-minded. Show how AI-generated media can move from prompt to pipeline to durable storage on Backblaze B2.
Get started
- Create a Backblaze B2 account: Backblaze B2 Signup
- Create a B2 bucket for your generated media, metadata, and app assets.
- Use the Genblaze SDK and star ⭐ the Genblaze GitHub Repository for examples, updates, and releases.
- Create a GMI Cloud account and submit the credits request form, if eligible. Use of GMI Cloud is optional, and other cloud providers may be used.
- Build and submit your generative AI media app on Devpost before the deadline.
The first 270 eligible participants will receive GMI Cloud credits to access generative AI and open source models for image, video, audio, chat, reasoning, and multimodal workflows.
To provide feedback on the Genblaze SDK, go to the Issues section of the Genblaze repo, click "New issue," and create a Bug report, Feature request, or start with a Blank issue.
Requirements
What to Build
Build a generative AI media application that uses Backblaze B2 and Genblaze.
Example project ideas include AI video generation or editing apps, image generation, remixing, or transformation tools, audio, music, or voice generation workflows, multimodal apps that combine text, image, video, and audio, AI media libraries for storing, organizing, and searching generated assets, provenance-aware workflows that track how media was generated, agentic media pipelines that generate, evaluate, retry, and store outputs, and creator, marketing, education, entertainment, or developer tools.
Projects should solve a real user problem and demonstrate a path toward production, including reliable media generation, durable storage, clear workflows, and a usable experience.
What to Submit
Submit your project through Devpost with:
- Working app: A functional app url judges can access and evaluate
- GitHub repo: Public or private code repository with setup instructions. Private repositories must grant judge contributor access for review.
- Providers and models: List the AI providers and models used
- B2 and Genblaze usage: Explain how your app uses both
- Demo video: Include a short (~3-min) demo that shows the Project functioning
Prizes
Grand Prize
Second Place
Third Place
Feedback prize
10 winning teams will receive a one-hour mentorship session with Backblaze engineers to review and optimize their Gen Media application architecture.
Devpost Achievements
Submitting to this hackathon could earn you:
Judges
Internal
Judging Criteria
-
Real-World Utility
Does the app solve a practical problem for a clear audience, and would that audience actually use it? -
Production Readiness
Does the app function reliably and support real-world workflows beyond a simple demo? -
B2 Storage and Data Orchestration
Does the app use Backblaze B2 meaningfully to store, organize, serve, or manage generated media, metadata, provenance, or app assets? -
Use of Genblaze
Does the app use Genblaze meaningfully to build, connect, or orchestrate generative media workflows across models, providers, or steps?
Questions? Email the hackathon manager
Invite others to compete
This site is protected by reCAPTCHA and the Google Privacy Policy and Terms of Service apply.


