The Lunabot model has been updated, and both GPT3 and GPT4 have adopted the latest 1106 model, which is faster and provides a better experience.
GPT-4 Turbo is even more powerful and is aware of world events up until April 2023. It has a 128k context window, so it can accommodate the equivalent of more than 300 pages of text in one prompt.
The new version of GPT-3.5 Turbo, this version by default supports a 16K context window. The new 3.5 Turbo supports improved instruction tracking, JSON mode, and parallel function calls. For example, our internal evaluations showed a 38% performance improvement when performing format tracking tasks (such as generating JSON, XML, and YAML).
We will release more updates in the future, so stay tuned.