Start Here
Four quick sections. You'll know exactly what MatsyaFlow is, whether it's for you, and what to do first.
What is MatsyaFlow?
MatsyaFlow provides free guides and tools to help developers cut their LLM API costs by routing tasks to the right model — cheap for simple work, powerful only when needed. Everything here is free: articles, tools, PDF guides, and a weekly newsletter.
What is OpenClaw?
OpenClaw is a self-hosted AI agent gateway. You define agents, skills, and how they interact — OpenClaw handles execution. It's open-source, runs on any hardware (including a Raspberry Pi), and lets you connect to any LLM via OpenRouter. These guides are built for OpenClaw users, but the routing principles apply to any direct LLM API setup.
Why does model routing matter?
Most AI setups send every request to the same model — usually the most expensive one. That's a problem: a task like 'summarise this paragraph' costs the same as 'analyse these 10 contracts'. With hybrid routing, you send simple tasks to a cheap model (DeepSeek, ~$0.28/M tokens) and only escalate complex tasks to frontier models (Claude Opus, ~$75/M tokens). The result: 70% cost reduction with no quality loss.
Is this for me?
Yes, if: you're running or building AI agents that make LLM API calls, you're paying for Claude, GPT-4o, or similar models, and you want to reduce spend without switching models entirely. No, if: you're using AI products (ChatGPT, Perplexity) rather than building with the API.