Applications using stable model names for GPT-3 base models (
davinci) will be automatically upgraded to the new models listed above on January 4, 2024. The new models will also be accessible in the coming weeks for early testing by specifying the following model names in API calls:
Developers using other, older completion models (such as
text-davinci-003) will need to manually upgrade their integration by January 4, 2024 by specifying
gpt-3.5-turbo-instruct in the “model” parameter of their API requests.
gpt-3.5-turbo-instruct is an InstructGPT style model, trained in the same way as
text-davinci-003. This new model is a direct replacement for the Completions API and will be available in the coming weeks for preliminary testing.
Developers wishing to continue using their optimized models beyond January 4, 2024 will need to tweak replacements on top of the new GPT-3 base models (
davinci-002), or more recent models (
gpt-4). Once this feature becomes available later this year, we will grant priority access to fine-tuning GPT-3.5 Turbo and GPT-4 to users who have already tweaked older models. We recognize that it is difficult to migrate from models refined on your own data. We will provide support to users who have previously tweaked their models to make this transition as smooth as possible.
In the coming weeks, we will contact developers who have recently used these older models and provide more information once the new finalized models are ready for initial testing.