AI Coding Field Notes

Coding agent pricing comparison: cheapest LLM API for coding, cost per million tokens, from the OpenRouter price list as JSON and CSV. An AI model token price comparison repriced at a real agent mix — 95.6% cached input pricing, 6.5x off list: prompt caching price, cache read price, cache hit share, long context pricing, DeepSeek peak/off-peak.

View the Project on GitHub xyzs996/llm-api-pricing

Revenue Growth

4 of the 49 write-ups here are tagged Revenue Growth. Every figure quoted below is in the figures table with the sentence it came from.

The figures

All figures, 408 rows

The write-ups

AI Local Websites Don’t Rent for $3K/Month—Until You Do This

You can build AI-generated local business websites, rent them to plumbers or dentists for $500–$3,000 a month, and scale to passive income.

reply box

Charge Per Conversation, Not Per Seat: The Billing Model Behind AI Support

Respond.io bills its customers for every contact who had a conversation in a given month rather than for every employee login, and I think that one decision explains more about the company’s positi…

reply box

One Person, 8 AI Agents, 3,000 Baseball Caps in Two Months

Zhang Qianchao runs a custom baseball cap export business on Alibaba.com with 8 AI agents and no employees, and in 2 months he shipped 3,000 caps to buyers across Europe, the Americas and South Ame…

reply box

Why Vanity Metrics Kill AI Startups: 700 Customers and 60,000 RMB From One Niche Account

In special-purpose vehicles — tankers, sweepers, the trucks nobody films for fun — one effective sales lead costs somewhere between 100 and 1,000 yuan, and a buyer who actually signs wires a hundre…

reply box


All 49 write-ups


Did this answer it? A star on the repository is the whole ask — it is what puts these in front of the next person looking; the data is CC BY and does not require starring.

Want a figure that is not here yet? Say which metric, which provider, which unit — one required field, and the page you came from is already filled in.

Got a better number? Open an issue — corrections and counter-data are the point.