GPT-4o vs Claude vs Gemini: Token Cost Comparison

Input & output pricing per 1M tokens — updated 2026

Every major AI provider charges by the token — and prices vary by more than 100× between the cheapest and most expensive models. Picking the right model for a task can cut your bill dramatically. Here's how the popular models compare.

Token pricing at a glance

ModelInput / 1MOutput / 1MContext
GPT-4o$2.50$10.00128K
GPT-4 Turbo$10.00$30.00128K
GPT-3.5 Turbo$0.50$1.5016K
Claude 3.5 Sonnet$3.00$15.00200K
Claude 3 Haiku$0.25$1.25200K
Gemini 1.5 Pro$3.50$10.501M
Gemini 1.5 Flash$0.075$0.301M
DeepSeek V3 / R1$0.27$1.10128K
Llama 3.1$0.59$0.79128K

Published list prices; actual costs vary by provider and volume discounts.

Key takeaways

Estimate your real cost

List prices only tell half the story — what matters is how many tokens your text uses. Paste your prompt into our token counter to see the exact count per model, or use the cost calculator to project a monthly bill from your request volume.

💰 Calculate your token cost

Related guides