Source: Pricing (OpenAI API)

OpenAI's per-token prices across Standard, Batch/Flex and Priority tiers, including cached input and long-context surcharges.

Contents
  • What: official API documentation page
  • Publisher: OpenAI
  • Date: accessed 2026-11-14
  • Original: Pricing

Summary

Gives input, cached-input and output prices for current and legacy models across processing tiers, plus surcharges above 272K input tokens.

Key claims

  • Batch and Flex are half of Standard; Priority processing was renamed Fast mode on 2026-07-30 and costs 2x Standard.
  • Cached input is 10x cheaper than fresh input on GPT-6 Astra ($1.00) and Luna ($0.01), and 20x cheaper on Sol ($0.10) and Terra ($0.05).
  • Requests above 272K input tokens pay higher rates (Astra $20/$75, Sol $4/$15, Terra $2/$7.50, Luna $0.20/$0.75).
  • Regional (data-residency) endpoints carry a 10% uplift for models released on or after 2026-03-05; FedRAMP endpoints also carry 10%.
  • Legacy models remain priced (GPT-5.x, GPT-4.1, GPT-4o, o3, o3-pro), useful for migration cost comparisons.

Feeds

Frontier model landscape · Model selection

Suggest changes →