On Friday, Google launched API pricing for Gemini 2.5 Professional, an AI reasoning mannequin with industry-leading efficiency on a number of benchmarks measuring coding, reasoning, and math.
For prompts as much as 200,000 tokens, Gemini 2.5 Professional prices $1.25 per million enter tokens (roughly 750,000 phrases, longer than the complete βLord of The Ringsβ sequence) and $10 per million output tokens. For prompts higher than 200,000 tokens (which most of Googleβs opponents donβt help), Gemini 2.5 Professional prices $2.50 per million enter tokens, and $15 per million output tokens.
That pricing makes Gemini 2.5 Professional dearer for builders than some other AI mannequin at the moment provided by Google, together with Gemini 2.0 Flash ($0.10/M enter tokens, $0.40/M output tokens). It additionally makes Gemini 2.5 Professional dearer than a number of different frontier AI fashions, akin to OpenAIβs o3-mini ($1.10/M enter tokens, $4.40/M output tokens) and DeepSeekβs R1 ($0.55/M enter tokens, $2.19/M output tokens).
To be honest, Gemini 2.5 Professional, which is out there free of charge with strict price limits, is available in cheaper than another extremely aggressive fashions, together with Anthropicβs Claude 3.7 Sonnet ($3/M enter tokens, $15/M output tokens) and OpenAIβs GPT-4.5 ($75/M enter tokens, $150/M output tokens). The tech {industry}βs preliminary response has been largely constructive, with builders praising what they understand to be smart charges.
However broadly talking, there appears to be some upward stress on pricing for flagship fashions. The price of current top-of-the-line releases from labs like Google, OpenAI, and Anthropic has been going up, not down. See, for instance, OpenAIβs just lately launched o1-pro, which is the corporateβs most costly API providing but at $150/M enter tokens and $600/M output tokens.
It may very well be that prime demand and computing prices are driving the development. In accordance with Google CEO Sundar Pichai, Gemini 2.5 Professional is the corporateβs most in-demand AI mannequin amongst builders, resulting in an 80% improve in utilization in Googleβs AI Studio platform and the Gemini API this month alone.
