Is There a Plan to Enable Function Call Support for the gpt-4-vision-preview Model?

The GPT-4-Vision-Preview model has shown immense promise in image processing and analysis. However, its inability to integrate function calls into its operations hinders its adaptability and integration into broader, more complex systems. This limitation is not just a minor inconvenience; it represents a significant barrier to leveraging the full potential of the model in various professional and innovative contexts.

The integration of function call support would transform how we can utilize this model, opening up new avenues for automation, efficiency, and creative application. The capacity to directly invoke model functions within varied operational contexts would significantly streamline processes and enhance the overall user experience.

I am fully aware of the technical complexities that such an integration might entail. However, I believe addressing this issue is essential for the model to remain competitive and relevant in rapidly evolving technological landscapes.

There is no plan stated. The status is “not currently”.

The clever programmer can shift model selection between models most suitable for the task at hand. The output of a specialized model does not need to be presented directly to the end user; instead the image description or data extracted can be provided to the primary AI model that will continue to interact with the user or can employ the image information for other purposes.

1 Like

see update here: