Frequently Asked Questions
Why are input tokens so much more expensive than expected?
Every time an AI coding agent modifies a file, it must re-read the existing codebase structure, the system prompt, tool definitions, and recent conversation history. As the project grows, this context compounds, making input costs 10–20× higher than output costs.
Which AI coding tool is cheapest to use?
Web UI copilots (ChatGPT, Claude.ai) are cheapest in API costs. For automated coding, CLI agents using budget models (GPT-4.1 nano, DeepSeek V3.2, Qwen3 30B) offer the best balance of automation and cost.
What is a turn in AI coding?
A turn is one complete request-response cycle with the AI model. CLI agents might execute 4–8× more turns than a web UI because they autonomously iterate — reading files, writing code, running tests, and fixing errors.
How accurate are these estimates?
Our estimates model the general pattern of context growth in AI-assisted coding. Real-world costs can vary by 30–50% depending on prompt engineering efficiency, caching, and project complexity.
Can I reduce AI coding costs with caching?
Yes. Anthropic’s prompt caching can reduce input costs by up to 90% for repeated context. With effective caching, your actual costs could be 40–60% lower for the input token portion.
Does model choice affect code quality?
Significantly. Premium models produce higher-quality code but cost 5–50× more than budget alternatives. Budget models work well for straightforward tasks but may require more iteration for complex logic.
Where does the pricing data come from?
All pricing is sourced from the OpenRouter API and verified against each provider’s official pricing pages. Prices are in USD per million tokens.
Are these prices per token?
The table shows prices per million tokens, which is how most LLM API providers publish pricing. The calculator converts your entered token counts into estimated dollar costs.
Does this include prompt caching?
The calculator uses standard input and output token costs. Some providers offer cache-read or batch discounts, so real bills can be lower when a workflow uses caching effectively.
How often are prices updated?
We verify and update model pricing weekly. New models are added as providers announce them. Check the “Last verified” date on the pricing page for the most recent update.
When should I use the pricing page vs. the homepage estimator?
Use the pricing page when you already know your input and output token counts. Use the homepage estimator when you want to estimate a full AI coding project from project size, tooling, features, and quality level.