🐱 dogecat

AI infrastructure
that just works

Memory, chatbots, and inference — everything you need to build AI-powered products. One platform. Simple APIs. Insanely cheap.

Three products. One platform.

Use them together or separately. Each product has its own API and dashboard.

🧠

Memory

Persistent, encrypted memory for AI agents. Semantic search across conversations.

  • • Client-side encryption
  • • Vector search
  • • Auto-linking memories
💬

Chatbots

Train chatbots on your docs. Embed anywhere with a simple script tag.

  • • RAG-powered answers
  • • PDF, URL, text sources
  • • Embeddable widget
⚡

Inference

Decentralized LLM inference. GPT-4 quality at a fraction of the cost.

  • • OpenAI-compatible API
  • • 10x cheaper
  • • No rate limits
10x
Cheaper inference
<100ms
Memory search
5 min
To deploy chatbot
99.9%
Uptime SLA

Simple APIs

# Store a memory
POST /api/v1/memories
{"content": "User prefers dark mode"}
# Create chatbot
POST /api/v1/chatbots
{"name": "Support Bot"}
# Run inference
POST /api/v1/try/completions
{"model": "gpt-4-turbo", ...}
# Search memories
GET /api/v1/search?q=preferences
→ [{content: "User prefers..."}]

Ready to build?

Free tier on all products. No credit card required.