A long journey. A whole toolkit.
Your first API call is only the beginning. As you build, you need to compare prices, understand usage, protect your margins, and decide what to run yourself. Narev builds free, open-source tools for that journey. From model prices to infrastructure costs, you can see what you spend and decide what to change. Start with the problem in front of you. Bring in the other tools when you need them.Every cost has a context
A model’s rate is one part of the story. How much you use it, who uses it, and where you run it matter too.
Pricing API
Compare before you commit. Look up provider rates and calculate inference costs through public, keyless REST or MCP. Give your usage numbers a price to work with.
AI-billing
Turn requests into cost visibility. Capture Vercel AI SDK usage, attribute it to customers, and send calculated events to your billing destination. See what serving your app actually costs.
Iwant
Explore another way to run inference. Deploy an open-weight model on a cloud GPU in your own account. Get an OpenAI-compatible endpoint and evaluate self-hosting for your workload.
Varne
Bring the infrastructure bill into view. Understand the infrastructure COGS (Cost of Goods Sold) behind your application, alongside model usage. Explore Varne v0; Varne v1 is in development.
IRL Gym
Keep quality in the equation. Explore real-world AI harness benchmarks to inform the tradeoffs you make alongside cost.