Sam Altman just announced OpenAI’s plans for GPT-4.5 and GPT-5, and they’re finally getting rid of that annoying model picker. Instead of making users guess which model to use, OpenAI is building unified systems that just work.
GPT-4.5, which they called Orion internally, will be their last model without chain-of-thought reasoning. After that, GPT-5 will combine all their tech into one system that knows when to think deeply or act quickly.
The really big news? The free version of ChatGPT will include unlimited access to GPT-5 at standard intelligence settings. Plus subscribers get a smarter version, and Pro subscribers get the highest intelligence level with extra features like voice, canvas, search, and deep research.
They’re also rolling their o3 model into GPT-5 instead of releasing it separately. This marks a clear shift away from making users pick between different models and toward AI that just does what you need.
This update makes a lot of sense. OpenAI built too many different models and complicated their product lineup. Now they’re fixing that by creating one system that handles everything – writing, coding, research, and analysis. I would like choice for devs and power users though.
For more context on OpenAI’s strategy shifts, check out my coverage of their Super Bowl ad campaign that cost them $14M: https://adam.holter.com/openai-spends-14m-on-super-bowl-ad-to-make-chatgpt-look-like-fire-and-wheels/
The timing and direction of these changes suggests OpenAI is focused on making their tech more accessible while maintaining different tiers of capability. By simplifying their offerings and providing free access to GPT-5, they’re positioning themselves to expand their user base significantly.
