OpenAI has been on a hot streak, unveiling new AI models and updates on a regular cadence. Last Thursday, OpenAI unveiled its latest model, GPT-4o mini, which, as the name implies, packs the power of GPT-4o into a smaller package.
Also: Two ways you can build custom AI assistants with GPT-4o – and one is free!
So, what can you expect? Your experience with GPT-4o mini shouldn’t differ vastly from that with GPT-4o, with the model performing close to its predecessor on popular benchmarks that test reasoning, math and coding proficiency, multimodal reasoning, and more.
Rather, the differences should be more noticeable when GPT-4o mini is compared with other small models — such as Gemini Flash, Claude Haiku, and GPT-3.5 Turbo — which GPT-4o mini repeatedly outperforms across the benchmarks.
If you are ready to find out for yourself, keep reading to learn the two easy ways to access GPT-4o mini today.