AI & Developer Tools

AI Model Price Tracker

Every major model's current API price in one table, plus a running changelog of cuts, hikes, and new releases. Last verified: July 2026.

Current prices (per 1M tokens, standard tier)

ModelProviderInputOutputTier
Gemini 2.5 Flash-LiteGoogle$0.10$0.40Budget
GPT-5.4 NanoOpenAI$0.20$1.25Budget
Gemini 2.5 FlashGoogle$0.30$2.50Budget
Claude Haiku 4.5Anthropic$1.00$5.00Fast
GPT-5.6 LunaOpenAI$1.00$6.00Fast
Gemini 3.5 FlashGoogle$1.50$9.00Fast
Gemini 3.1 ProGoogle$2.00$12.00Flagship
GPT-5.4OpenAI$2.50$15.00Flagship
GPT-5.6 TerraOpenAI$2.50$15.00Flagship
Claude Sonnet 5Anthropic$3.00*$15.00*Flagship
Claude Opus 4.8Anthropic$5.00$25.00Frontier
GPT-5.5OpenAI$5.00$30.00Frontier
GPT-5.6 SolOpenAI$5.00$30.00Frontier
Claude Fable 5Anthropic$10.00$50.00Frontier

*Claude Sonnet 5 has an introductory rate of $2.00 / $10.00 through August 31, 2026 — list price shown. Estimate your own workload with the AI cost calculator.

Changelog

July 2026 — tracker launched (baseline)

The one trend to know

The price of a given level of AI capability has been falling fast — industry analyses put it around 10× cheaper per year for equivalent quality, driven by hardware, efficiency gains, and competition. Practically: the flagship model you priced six months ago probably has a successor that's cheaper and better. Re-check before committing an annual budget, and design your product so the model can be swapped easily.

How to read AI pricing pages without getting burned

Frequently asked questions

How often is this updated?

Monthly, and after any major provider announcement. The "last verified" date at the top tells you the snapshot's freshness.

Why do some sites show different prices?

Usually one of: outdated data, batch/cached rates presented as headline prices, reseller markups, or regional/enterprise tiers. We track the providers' standard public list prices.

Do prices differ by region?

List prices are global in USD; cloud-marketplace versions (AWS Bedrock, Google Vertex, Azure) can differ slightly and add enterprise features.