On Friday, Google was released API pricing For Gemini 2.5 Pro, an AI reasoning model with top performance at various reference points that measure coding, reasoning and mathematics.
For up to 200,000 chips, the Gemini 2.5 Pro costs $ 1.25 per million input brands (about 750,000 words, more than the entire “Lord of the Rings” series) and $ 10 per million brands. For prompts greater than 200,000 chips (which most of the Google competitors do not support), the Gemini 2.5 Pro costs $ 2.50 per million input brands and $ 15 per million tokens.
This pricing makes Gemini 2.5 Pro more expensive for developers than any other AI model offered by Google today, including Gemini 2.0 Flash (0.10/million input brands, $ 0.40/million brands). It also makes Gemini 2.5 Pro more expensive than many other AI Frontier models, such as O3-MINI ($ 1.10/million input brands, $ 4.40/million output brands) and Deepseek R1 ($ 0.55/million input brands, $ 2.19/Million tokens).
To be fair, the 2.5 Pro Gemini, which is available free of charge with strict interest rate limits, comes cheaper than some other highly competitive models, including Claude 3.7 Sonnet ($ 3/Millions of brands, $ 150/Millionai’s tokens tokens ($ 75/Millionai) and $ 150 Millions. was largely positivewith developers praising what they realize is a reasonable pace.
But in general, there seems to be some upward pressure on the pricing for flagship models. The cost of the recent releases of top lines from laboratories such as Google, Openai and Anthropic were increasing, not down. See, for example, the recently launched O1-PRO, which is the most expensive API of the company that still offers $ 150/million input brands and $ 600/million chips.
It could be that the high demand cost and computers leads the trend. According to Google Sundar Pichai’s CEO, Gemini 2.5 Pro is the company’s most AI model, leading to 80% increase in use on Google’s AI Studio Platform and API Gemini Only this month.
