Seats & Tokens

Home / Methodology

About & methodology

Seats & Tokens is an independent tracker of what AI coding tools cost. We publish data, not opinions dressed as data: every price links to the page we read it from and the date we read it.

What we track

17 tools (IDEs, terminal agents, app builders) and 28 models. Current snapshot: Oct 21, 2026.

Where prices come from

How data is refreshed

How the calculator estimates cost

  1. Your sessions per week × hours per session × model requests per hour × 52/12 weeks = requests per month.
  2. Requests × input/output tokens per request (with a cached-input share) × the model's per-token price = the API-equivalent cost.
  3. Each plan applies its published rule:
    • Dollar allowance: Copilot, JetBrains, Zed, Augment. We add overage, with markups where published (Zed +10%, Augment +40%).
    • Plan plus API-priced usage: Cursor for third-party models.
    • Flat price with an unpublished quota: Claude, Codex, Windsurf.
    • Vendor credits we can't convert: Kiro, Lovable, Bolt, Replit, v0. We show the plan price as a minimum.

The presets are editorial assumptions. They're visible and editable in the calculator.

Independence

No paid rankings. Verdicts are written from the published data. Sponsored content, if any, is labelled "Sponsored" and kept separate from data tables. See the disclosure.

Corrections

Found a wrong price? Corrections are logged publicly on the change log. A corrections email address will be published here at launch.

Open data

/api/prices.json, /api/models.json, /llms.txt and /llms-full.txt. Licensed CC BY 4.0.