o4-mini Pricing

o4-mini API pricing

o4-mini API pricing is $1.1 per million input tokens and $4.4 per million output tokens, verified November 1, 2026.

Input (per 1M tokens)

$1.1

Output (per 1M tokens)

$4.4

Cached input (per 1M tokens)

$0.28

Context window

200,000 tokens

Max output

100,000 tokens

Capabilities

o4-mini supports extended reasoning, image input, tool use.

o4-mini API pricing has changed 2 times on record, most recently on October 12, 2026 (increase).

• October 12, 2026: Increase — input $0.55 → $1.1/M (+$0.55) — output $2.2 → $4.4/M (+$2.2)
• August 24, 2026: Decrease — input $1.1 → $0.55/M (−$0.55) — output $4.4 → $2.2/M (−$2.2)

Frequently asked questions

Short answers to the most common questions about o4-mini pricing, limits, and capabilities.

Notes

Long Context flag corrected to false 2026-07-28 — o4-mini has 200K token context, not 1M+. [2026-08-24] PRICE CUT — Input fell $1.10 → $0.55 and Output fell $4.40 → $2.20 per million tokens (50% cut on both sides). Corroborated by cloudzero.com "OpenAI API pricing in 2026: every model after the July price cuts", pricepertoken.com o4-mini page ("cheapest input price has fallen 50%, from $1.10 to $0.550 over the past ~90 days"), and metacto.com. Cut appears tied to OpenAI's July 30, 2026 price reductions. [2026-10-09] Verification attempt produced no clear result — recheck next run. This run's sources (anotherwrapper dated 2026-08-25 citing OpenAI's page; requesty; llmreference) all list $1.10/$4.40 standard, and one lists $0.55/$2.20 as the BATCH (-50%) rate. Risk: the stored $0.55/$2.20 may be the batch price, not a standard-rate cut. No primary read obtained. Prices LEFT at $0.55/$2.20 pending a direct read of openai.com/api/pricing. [2026-10-10] Verification attempt produced no clear result — recheck next run. Second consecutive run where aggregators (anotherwrapper, pricepertoken) list o4-mini at $1.10/$4.40 standard with $0.275 cached; none this run reproduces $0.55/$2.20 as a standard rate. Strong signal the stored values are the batch rate, but no primary openai.com read — prices LEFT at $0.55/$2.20. ACTION FOR KHALED: confirm on openai.com/api/pricing and revert to $1.10/$4.40 if confirmed. [2026-10-11] Verification attempt produced no clear result — recheck next run. Sources still split: one (May 2026 metacto) gives $0.55/$2.20, another gives $1.10/$4.40. No primary openai.com read. Prices LEFT at $0.55/$2.20. Third consecutive conflicted run. [2026-10-12] DATA CORRECTION — Input $0.55 → $1.10, Output $2.20 → $4.40, Cache Read (empty) → $0.275. OpenAI's own model page (platform.openai.com/docs/models/o4-mini) lists Input $1.10 / Cached $0.275 / Output $4.40; corroborated by anotherwrapper (citing OpenAI), pricepertoken and an Oct 8 2026 aggregator table. The 2026-08-24 'cut' was the 50% Batch rate recorded as standard. Snapshot written (Increase) so the calculator change is traceable, but this is a correction, not a vendor price rise. ALSO: OpenAI's model page now marks o4-mini as DEPRECATED (successor: GPT-5 mini) — Khaled to decide on Status.