[ May 31 - Applied Intelligence Hackathon ]
Hackers may receive $5.00 worth of AI credits on cloud.near.ai, which may be used to cover private inference through NEAR AI deployed models. We recommend GLM-5 or Qwen 3.5.
Claiming NEAR AI Cloud Credits
<aside> 🏆
NEAR is a high-performance, sharded layer-1 blockchain purpose-built to support AI-native and user-owned Web3 applications at internet scale.
With a design focused on real-time, low-cost execution and advanced features such as encrypted model execution, chain abstraction, and autonomous AI agents, NEAR aims to empower builders with scalable infrastructure for both AI and decentralized applications.
https://docs.near.org/getting-started/hackathons
We’ve prepared suggestions for NEAR integrations with each of the sponsors:
Build an application for Healthcare, Law, Finance, etc that leverages NEAR AI’s private inference for secure conversation, and then store these sensitive dataset assets in Tigris Data.
Deploy an IronClaw agent outfitted with a specialized skill to autonomously provision Opsera’s DevSecOps Agents, automating security scanning and code hardening for the Cybersecurity track.
Integrate Insforge’s unified agent-optimized backend with IronClaw via a custom skill to build agents that autonomously trigger databases, custom functions, and auth to complete transactions end-to-end via NEAR Intents.
Build a Marketing or Research & Science agent that gathers web intelligence via Apify's extraction workflows and analyzes it using NEAR AI private inference to prevent competitive scraping data from leaking, or executes multi-chain transactions via NEAR Intents.
Orchestrate multi-agent software factories using Lightsprint.ai to build AI & Tech tools that dynamically execute off-chain task logic and trigger settlement via NEAR Intents.
Scale your generative enterprise workloads on Nebius's top-tier GPU infrastructure while ensuring total data confidentiality through NEAR AI (Private Inference) powered by TEE-based secure enclave enforcements.
Leverage Daytona's isolated sandboxes to use AI agents like IronClaw or OpenCode to build full-stack applications with via the everything-dev framework.
Utilize Kalibr’s orchestration layer to coordinate multimodal Art and Creativity agents across different generative models while preserving creator data ownership via NEAR AI private inference.
Secure personal data, meeting transcripts, and raw notes captured by Brain2 for the Education or Research tracks by processing all highly private mental digests through **NEAR AI private inference.**
Incorporate Rtrvr.ai's browser-automation capabilities as an IronClaw skill to build agents that autonomously execute complex web tasks and API actions without exposing active user session cookies.
Deploy your production-grade applications built with NEAR Intents and NEAR private inference directly on Render for reliable, auto-scaling deployment.
</aside>
<aside> ❓
First, consider using one of the https://github.com/near/agent-skills or finding a related resource in our hackathon starter kit: https://docs.near.org/getting-started/hackathons
Then, further explore our documentation:
And then if you still can’t find your answer, we have lots of developer resources available to answer questions in several telegram chats:
</aside>