OpenAI Launches o3 Pro But What's New About It?

OpenAI dropped o3 pricing by 80% and launched o3 Pro today, positioning it as the most powerful reasoning model in their lineup. The base o3 now costs $2 per million input tokens and $8 per million output tokens. Meanwhile, o3 pro carries a $20/$80 rate with access through both ChatGPT Pro/Team and the OpenAI API.

In controlled tests, human reviewers preferred o3 Pro for clarity, accuracy, and follow-through, particularly on science, coding, education, and business puzzles. Benchmarks such as AIME 2024 and GPQA Diamond show o3 pro outperforming Google’s Gemini 2.5 Pro and Anthropic’s Claude Opus on advanced math and science.

Early user feedback reveals a clear pattern: o3 pro outperforms when fed deep context. Raindrop.ai’s Ben Hylak collected meeting notes, goals, and voice memos before prompting o3?pro. The model then generated a concrete roadmap with deadlines, priorities, and experimental cut-offs that reshaped the team’s strategy.

Other recent Open Ai news –

• OpenAI Lanches Data Residency Options for Enterprise and Educational Users
• OpenAI’s o3 and o4-mini Models Can Now Analyze Images Like a Human
• Google opens AI campus in London & & expands learning for young girls nationwide

Users say o3 pro sets itself apart by reasoning about tool use. It recognizes when it lacks external info. It asks to fetch data or run code. It asks probing questions. That ability places it a step ahead of its base model. But small-context tasks sometimes stifle it. Some queries that worked better on o3 or Claude returned weaker results under o3 pro’s heavier compute budget.

Reddit discussions reflect this. One early reviewer noted that o3 pro fails easily on visual tasks and hallucinations still occur without enough direction . Others report that while benchmarks appear saturated, practical code and project planning tasks show real-world improvements .

The AI giant’s move signals a shift toward context-aware, tool-smart reasoning engines. For users ready to build memory-rich agents or heavyweight planning flows, o3 pro delivers real gains. If your tasks stay light and chatty, stick with the lighter, nimbler options.

You may also be interested to read –

• Three major newsrooms roll out AI-based bullet-point summaries
• Getty takes Stability AI to UK court over photo copyrights
• China Clamps Down on AI During High-Stakes Gaokao Exams