GPT-4o mini: prices collapse, 'good enough' AI becomes nearly free
In one sentence OpenAI ships GPT-4o mini at $0.15/$0.60 per 1M tokens, 60% cheaper than GPT-3.5 Turbo, MMLU 82%. Moves the 'baseline model' bar for most use cases.
Since the OpenAI API has existed, the default "cheap model" has been GPT-3.5 Turbo. Cheap, fast, but clearly dumber than GPT-4.
GPT-4o mini replaces it entirely: 60% cheaper and better on almost every benchmark. $0.15 per million input tokens, $0.60 per million output — meaning an app can summarize a novel for pennies.
For AI product builders it's a turning point: the whole category of tasks "don't need genius, need a decent assistant" — classification, extraction, formatting, translation, summarization — can be done at near-zero cost. You can run agent loops with dozens of calls without watching the bill explode.
Knock-on effects: Anthropic cuts Haiku prices, Google trims Gemini Flash, Meta accelerates Llama 3.1.
Companies
OpenAI
Tools
GPT-4o mini
Tags
Sources