Back in April, OpenAI released o3, its flagship reasoning model, which delivered state-of-the-art (SOTA) results on several AI benchmarks, including Codeforces, SWE-bench, and MMMU. The o3 model also included support for external tools, such as web browsing, a Python interpreter, and more.
Although o3 was the most powerful model available from OpenAI, it was not widely adopted by developers due to its pricing compared to other alternatives. Google’s Gemini 2.5 Pro model offered nearly the same performance at one-quarter of the cost, while Anthropic’s Claude 3.7 models delivered better coding performance at a similar price.
Today, OpenAI is responding to developer feedback by reducing the price of the o3 model by nearly 80%. At launch, the o3 model was priced at $10 per million input tokens and $40 per million output tokens. As of today, OpenAI is lowering the price to $2 per million input tokens and $8 per million output tokens. This price drop makes o3 a much more attractive option for developers.
we dropped the price of o3 by 80%!!
— Sam Altman (@sama) June 10, 2025
excited to see what people will do with it now.
think you'll also be happy with o3-pro pricing for the performance :)
You can refer to the benchmark comparison table below, which Google published last week during the launch of the Gemini 2.5 Pro June update, to see how OpenAI’s o3 compares with other leading models.
In addition to the price reduction, OpenAI is also launching o3 Pro, its new flagship reasoning model, later today. The o3 Pro model is expected to outperform the Gemini 2.5 Pro June update on several key AI benchmarks and may be priced at $10 per million input tokens and $40 per million output tokens.
6 Comments - Add comment