OpenAI's new models are now on GPTunneL: meet o3 and o4-mini!

OpenAI's new models are now on GPTunneL: meet o3 and o4-mini!

Great news! GPTunneL just got a new addition: we've added two of OpenAI's freshest models — o3 and o4-mini. Now you have even more great tools for any task, from deep analysis to fast on-the-fly answers.

What can OpenAI's new models do?

OpenAI has presented o3 and o4-mini as a major step forward in its technology. Each model has its own strengths, backed by benchmark results.

OpenAI o3: power for complex tasks

This model is described as OpenAI's strongest tool for deep reasoning. It excels wherever precision and analysis matter. O3 sets new benchmark records, outperforming previous models, especially on complex tasks.

o3 particularly shines at:

  • Coding: delivers top results on Codeforces (2706 points) and SWE-bench.
  • Math and science: scores 98.4% on AIME 2025 using Python and 83.3% on GPQA. In practice, this means the model handles PhD-level problems with ease.
  • Visual perception: deeply analyzes images and charts, even when blurry, rotated, or low quality.
  • Query accuracy: makes 20% fewer major errors than o1, according to expert evaluations.

O3 is a great assistant for developers, researchers, and anyone working with complex information who values precision.

OpenAI o4-mini: speed and efficiency

This model is a compact, fast option. The main advantage of o4-mini is speed and cost-efficiency combined with impressive performance for its size. It's built for situations where you need an answer fast without sacrificing quality. O4-mini often outperforms even its predecessor, o3-mini.

What makes o4-mini great:

  • Fast, efficient answers. The model can reason through your queries quickly and deliver detailed responses.
  • High performance in math and code. Scores 99.5% on the American Invitational Mathematics Examination (AIME 2025) using Python, and 2719 points on Codeforces — among the highest scores of any AI model.
  • Content creation and data analysis. Great for quickly writing articles, summarizing text, classifying data, and answering emails.
  • Lower price than o3. Check generation costs on our pricing page.

The model is smart enough for most everyday tasks while remaining very fast and affordable. With o3 and o4-mini, OpenAI offers a choice: maximum power for demanding challenges, and fast, economical solutions for everyday needs.

Where can you try o3 and o4-mini?

Want to put the new models to the test? Nothing could be easier!

GPTunneL gives you direct access to o3 and o4-mini. Forget about VPNs or complicated setup. Just pick the model in our interface and start creating! Discover new possibilities with OpenAI's cutting-edge neural networks on GPTunneL!