Calculators, comparisons and setup guides that run on the same pricing and leaderboard data.
Two or more models in one table, optionally hiding rows where they agree.
Enter monthly usage and a cache hit rate to rank monthly cost by channel.
Token counts and cost for one piece of text under several tokenizers.
Weigh a subscription against pay-as-you-go API at your usage level.
Plan prices and quotas for ChatGPT, Claude, Gemini and other consumer subscriptions.
Pricing and leaderboard data as free static JSON, with examples for code and MCP-enabled agents.
Point Claude Code, Codex CLI, Gemini CLI, Aider, opencode and Qwen Code at your own API key and endpoint.
The calculators run on the same pricing and leaderboard data as the rest of the site. The cost calculator prices your monthly token usage on both the official and cheapest channel, with a separate cache-hit input because cached input is billed well below the normal input rate.
The token counter shows how the same text tokenizes differently across model families, which matters most for Chinese and other CJK text where newer tokenizers need fewer tokens per character. The subscription comparison weighs a flat monthly plan against pay-as-you-go API at your usage level, reported as a range rather than a single break-even point.