OpenAI API Pricing 2026

OpenAI remains the largest AI API provider in 2026, offering models across four pricing tiers — from the budget-friendly GPT-4.1 nano at $0.10 per million input tokens to the premium o1 reasoning model at $15.00 per million input tokens.

The biggest change in 2026 was the launch of the GPT-5.4 family, which brought significant capability improvements while maintaining competitive pricing. GPT-5.4 costs $2.50/M input tokens — the same as GPT-4o — but with substantially better reasoning, coding, and instruction-following abilities.

GPT-5.4 Family

The GPT-5.4 lineup covers three tiers:

• GPT-5.4 ($2.50 input / $15.00 output) — Flagship model with 270K context. Best for complex reasoning, creative writing, and nuanced tasks.
• GPT-5.4 mini ($0.75 input / $4.50 output) — Great balance of cost and capability. Ideal for most production workloads.
• GPT-5.4 nano ($0.20 input / $1.25 output) — Budget option for high-volume tasks like classification, extraction, and routing.

o-Series Reasoning Models

OpenAI's reasoning models use internal "thinking tokens" that increase actual costs beyond the listed per-token rates. The o3 model ($2.00/M input) replaced the expensive o1 ($15.00/M) as the primary reasoning model — an 87% price cut with improved performance.

o4-mini at $1.10/M input tokens offers the most cost-effective reasoning capability in OpenAI's lineup.

GPT-4.1 Family

Released alongside GPT-5.4, the GPT-4.1 family targets developers who need a 1M-token context window:

• GPT-4.1 ($2.00 input / $8.00 output) — 1M context, strong for long-document processing
• GPT-4.1 mini ($0.40 input / $1.60 output) — Best value for large context needs
• GPT-4.1 nano ($0.10 input / $0.40 output) — Cheapest model in OpenAI's lineup

Cached Input Pricing

All GPT-5.4 and GPT-4.1 models support prompt caching, which can reduce input costs by 75-90%. For example, GPT-5.4's cached input rate is just $0.25/M tokens — 90% cheaper than the standard $2.50/M rate.

How OpenAI Compares

OpenAI generally offers more aggressive pricing than Anthropic across all tiers. GPT-5.4 mini ($0.75 input) is 4x cheaper than Claude Sonnet 4.6 ($3.00 input). However, DeepSeek undercuts everyone at $0.28/M input for their flagship model.

For budget use cases, GPT-4.1 nano ($0.10/M) and GPT-4o mini ($0.15/M) compete directly with Google's Gemini 2.5 Flash-Lite ($0.10/M).