Filters

AI Models

Data

Redis Qdrant

Platforms

Streamlit Replit Vercel

Upcoming AI Hackathons
For Innovators & Creators

Discover AI Applications

Browse applications built on modern technologies. Explore PoC and MVP applications created by our community and discover innovative use cases for modern technologies.

TokenEdge — Hybrid AI Routing Agent

TokenEdge is a verification-gated AI routing agent that solves tasks locally whenever possible and uses Fireworks AI only when necessary, minimizing scored tokens while maintaining accuracy.

nexico

AMD Developer CloudGemmaQwen3+5

EcoRoute: 3-Tier Agentic Router

EcoRoute is a hybrid AI routing engine that intercepts easy tasks using a zero-token local model, dynamically escalating complex queries to the cloud. It slashes API costs while ensuring enterprise-grade resilience.

jhon jones

Llama 3

Cheapskate Router

A containerized AI agent that routes tasks across 8 capability categories to the cheapest sufficient Fireworks AI model, using zero-token local classification and syntax validation to minimize token spend while clearing the accuracy gate.

Crazybell

AMD Developer CloudAnthropic ClaudeAntigravity+5

Zero-Token Hybrid Routing Agent

A two-stage, hybrid routing architecture that intercepts queries using an optimized regex engine and a localized quantized GGUF model, escalating only complex logic and math tasks to the Fireworks API to maximize accuracy while minimizing token costs.

CuriousVJ

AntigravityDeepSeek R1Gemma

Local-First Token-Efficient Routing Agent

A hybrid routing agent that answers most tasks with a free local model and calls the cheapest sufficient Fireworks model only when needed — pushing token usage toward zero while clearing the accuracy gate across all eight Track-1 task categories.

Nural Nexus

Claude CodeAnthropic ClaudeAMD Developer Cloud+1

CipherRoute

CipherRoute is a token efficient hybrid AI agent that intelligently routes every task between deterministic local solvers and Fireworks models, delivering high accuracy while minimizing API usage, cost, and latency for general purpose AI workloads.

Opter

AMD Developer CloudCodexChatGPT+1

BenchSieve

BenchSieve is a format-aware AI task router that reduces Fireworks token usage by solving safe prompts locally and tightly budgeting remote model calls.

MYSELF

GPT-5OpenAIGemini AI+3

GemmaRoute: The Financial Firewall for AI

GemmaRoute is an intelligent semantic routing layer for enterprises. It logic-gates prompts, routing simple queries to $0-cost local AMD hardware (Gemma 2), while escalating complex reasoning to cloud models (Llama 3.1) to slash inference costs.

Gemma Labs

StreamlitAntigravityVercel+8

ROCm-Pilot: AI-Powered AMD Setup Assistant

ROCm-Pilot is an AI assistant that simplifies setting up machine learning environments on AMD hardware. It auto-detects system specs, searches official ROCm documentation, and generates tailored, copy-pasteable setup instructions with verifiable sources.

SoulsLikeAMD

AMD ROCmAMD Developer CloudLangChain+1

Zero-Cost Multi-Agent Token Router

A hybrid routing agent that minimizes API costs by classifying queries on the fly, leveraging fast local models for trivial tasks, and dynamically spinning up a local multi-agent debate committee for complex problems.

The Goods

Antigravity

CrisisCommand

Dashboards show you the crisis. CrisisCommand simulates your decision: live disaster feeds land on a holographic 3D globe, then a GPU Monte Carlo engine and open LLMs forecast escalation and present three response options with honest ranges.

Imitasi

AMD ROCm

Token-Router: Hybrid Token-Efficient Routing Agent

An autonomous agent that answers each task with the cheapest model that can get it right. Try a weak local model first, verifying its answer for free, and escalating to Fireworks only when it has to, to minimize total tokens without sacrificing accuracy.

JoyInAI

OpenAIAMD Developer Cloudrest api

routerx-agent

A smart-routing AI agent that handles 8 task categories using local models (zero cost) and Fireworks AI API (minimal tokens).

rm -rf

AntigravityQwen3

Token Frontier Router

A lightweight routing agent that cuts token usage by sending each task to the cheapest model that can still answer well.

Paripakv

Anthropic ClaudeAMD Developer CloudAI/ML API+1

Track 1 - Smart Router

Cost-aware routing agent: deterministic solvers and a bundled 3B local LLM answer most tasks at zero token cost; only genuinely hard tasks escalate to Fireworks AI models, minimizing billed tokens while clearing the accuracy gate.

Robot ROCm

Qwen3

REPOMIND v3 - on-prem AI coding agent on AMD

On-prem, air-gapped AI coding agent + LLM cost-router for enterprises barred from cloud AI - 100% AMD, open-weight, Ed25519 tamper-evident audit. $0/token on-prem; cuts existing token bills 40–70%. AMD ACT I 1st-place winner.

REPOMIND v3

AMD Developer CloudAMD ROCmGemma+1

Hybrid-Router

An intelligent LLM router that selects the most token-efficient Fireworks AI model for each task, optimizing accuracy and inference cost.

Code

AMD Developer CloudAMD ROCm

Token-Slasher: Minimalist Hybrid Routing Agent

A lean, containerized routing agent that maximizes token efficiency by automatically selecting small-footprint open models and stripping out conversational fluff.

Solo leveler

Llama 3

Hybrid Token-Efficient Routing Agent

An advanced dual-model orchestration pipeline combining a custom 30M Mini-BERT router, a local quantized INT8 Qwen-0.5B execution engine, and automated Fireworks API fallback escalation for token-efficient, zero-hallucination edge intelligence.

Absolut Void Development

AgentOpsAI/ML APIGemini 3 Flash+1

Hybrid Token-Efficient Router

An AI agent that answers 8 different task types by routing each one to a free local model when it can, and to Fireworks only when the task genuinely needs a stronger model, cutting token usage without hurting accuracy.

Jahnvi Singh

AMD Developer CloudLlama 3.1HuggingFace Hub

OptiRouter

Maximize accuracy and minimize API cost by dynamically routing queries to zero-token local inference with speculative remote correction.

AgentSmiths

Gemma 2AntigravityClaude Code+2

Token-Efficient Hybrid AI Gatekeeper

An intelligent, Dockerized hybrid AI proxy gateway that dynamically routes prompts between local Ollama edge nodes and Fireworks AI cloud tiers based on task complexity to minimize token spend.

LifeInvaders

AMD Developer CloudLlama 4AI/ML API+1

Axiom - Aether

Axiom is a local-first Rust research runtime for online Test-Time Training (TTT), context compression, grounding checks, and self-healing execution loops.

B Team

AI/ML APIAMD Developer CloudAMD ROCm+6

Dual Pitcher AI Agent

A hybrid AI Agent that can handle variety of natural language tasks across multiple disciplines, using either local or Fireworks AI models as efficiently as possible.

NPA

AI/ML APIAMD Developer CloudClaude Code+5

SatyaPath Verify: AI Welfare Claims Auditor

AI ensemble + SHAP + Gemma flag and explain fraudulent MGNREGA welfare claims in plain language, then cryptographically commit verified batches via Merkle proofs, built on AMD Developer Cloud and Fireworks AI.

solo

AMD Developer CloudGemmaAI/ML API+3

Marrow: Where FinOps Meets SecOps

An AI-powered command center correlating AWS Cost Explorer data with Security Hub findings. Powered by DeepSeek via Fireworks AI on AMD Instinct MI300X, Marrow optimizes cloud spend and reduces attack surfaces simultaneously.

Rikora

AMD Developer CloudAWSrest api+2

ApexFlow AI: Self-Healing Telemetry Gateway

ApexFlow AI is a self-healing edge telemetry gateway that uses a hybrid quantum-classical VQC router to autonomously reconcile schema drift in real-time, preventing mission-critical IoT data blackouts on AMD Instinct GPUs.

Tarek Clarke

AMD ROCmGemmaNVIDIA+2

Task_Blowing_Agent

An optimized wind like light, containerized AI agent with intelligent local task routing and server less API integration for efficient, token-saving execution.

Neural Knights

Gemini AINVIDIAAMD Developer Cloud

V3 Zero Token Hybrid AI Router

A highly efficient, cost-optimized AI agent that maximizes local zero-token processing via spaCy and quantized Llama 3.2, exclusively reserving Fireworks API compute for complex reasoning tasks to ensure extreme token efficiency and precision.

MMPC_solo

HuggingFace HubLlama 3.2

AI Crisis Command & Coordination Platform

An AI Crisis Command & Coordination Platform is a centralized, multi-agent operational operating system designed to structure chaos and synchronize real-time emergency responses during high-stakes incidents.

BAND OF HACKATHON

AntigravityGemini 3 FlashBright Data Web Scraper API

30apps loaded