OpenAI has announced the launch of its most advanced reasoning model yet o3 Pro for ChatGPT Pro and Team users. The new release is accompanied by a major price reduction of 80% on the standard o3 model, now priced at $2 per million input tokens and $8 for output tokens. OpenAI CEO Sam Altman confirmed the price cut and hinted that o3 Pro pricing will also reflect its strong performance-to-cost ratio.
The o3 Pro model replaces the earlier o1 Pro in ChatGPT and delivers highly reliable responses in domains such as mathematics, programming, science, and education. Optimized for depth and accuracy, o3 Pro includes access to advanced tools like web browsing, Python code execution, visual input analysis, and document handling. However, it currently lacks image generation capabilities and Canvas integration due to technical limitations. Temporary chats are also disabled for now.
Despite the momentum with o3 Pro, OpenAI has pushed back the launch of its long-anticipated open-source AI model, originally expected in June 2025. According to Altman, the delay stems from a recent breakthrough by the research team that requires additional time to finalize. While no new release date has been confirmed, Altman suggested it will arrive later this summer.
The open-source model is expected to challenge competitors like DeepSeek R1, aiming to set a new standard for publicly available reasoning AI. In the meantime, OpenAI’s new o3 Pro is being rolled out to enterprise and education users next week and positions the company to compete aggressively with models from Anthropic and Google DeepMind.
Additionally, OpenAI has addressed concerns around the environmental cost of AI queries. A single ChatGPT query uses about 0.34 watt-hours of electricity and 0.000085 gallons of water, according to Altman. The company’s long-term goal, he stated, is to drive the cost of intelligence closer to that of electricity. With o3 Pro setting new benchmarks for reliability and cost-efficiency, and a high-performance open-source model on the way, OpenAI continues to reshape the landscape of generative AI.
