OpenAI API Pricing 2026: Complete Guide to GPT-5, GPT-4.1, o3, and o4 Costs

OpenAI's pricing changes frequently, and keeping up with the latest costs for GPT-5, GPT-4.1, o3, o4, and other models can be challenging. Here's everything you need to know about OpenAI API pricing in 2026.

This guide covers current pricing for all OpenAI models, helps you estimate costs for your use case, and shows you how to track spending in production.

OpenAI API Pricing Overview (January 2026)

OpenAI charges per token—small units of text (roughly 4 characters or 0.75 words). Prices differ for input tokens (what you send) and output tokens (what the model generates).

GPT-5 Series (Latest)

GPT-5.2 is OpenAI's latest flagship—even faster and more capable than GPT-5, with improved reasoning and multimodal abilities.

GPT-5-mini and GPT-5-nano offer exceptional value for simpler tasks at a fraction of the cost.

GPT-4.1 Series

GPT-4o Series

Reasoning Models (o-Series)

The o-series models are designed for complex reasoning tasks—math, coding, and multi-step problem solving. They're more expensive because they "think" before responding.

Legacy Models

These models are still available but mostly superseded. GPT-4o is cheaper and better than GPT-4 Turbo. GPT-4o-mini is cheaper and often better than GPT-3.5 Turbo.

Real Cost Examples

Let's calculate costs for common use cases:

Customer Support Chatbot

Average conversation: 500 input tokens, 300 output tokens
Daily volume: 1,000 conversations

Monthly cost with GPT-4o: $165
Monthly cost with GPT-4o-mini: $10

Document Summarization

Average document: 4,000 input tokens, 500 output tokens
Daily volume: 500 documents

Code Generation

Average request: 800 input tokens, 1,200 output tokens
Daily volume: 200 requests

How to Reduce OpenAI API Costs

1. Use GPT-4o-mini for simple tasks — Classification, extraction, and basic Q&A work well with the smaller model
2. Optimize prompts — Shorter prompts = fewer tokens = lower costs
3. Cache responses — Don't pay twice for the same query
4. Set spending limits — OpenAI's dashboard lets you set monthly caps
5. Track costs by feature — Know which parts of your app cost the most

Tracking OpenAI Costs in Production

OpenAI's dashboard shows total usage, but not which features drive costs. To optimize effectively, you need per-feature tracking.

import { Orbit } from '@with-orbit/sdk';
import OpenAI from 'openai';

const orbit = new Orbit({ apiKey: process.env.ORBIT_API_KEY });

// Track costs by feature
const chatClient = orbit.wrapOpenAI(new OpenAI(), {
  feature: 'customer-chat'
});

const summaryClient = orbit.wrapOpenAI(new OpenAI(), {
  feature: 'doc-summary'
});

Now you can see exactly how much each feature costs and optimize where it matters.