I know openai has had vision models for a while now. Is there one that I can access through the api with either rest or websocket?
#Got sight?
wraith musk
empty turtleBOT
Hi @wraith musk. I'm kapa.ai, your friendly support bot. I will try to answer your question by searching through the knowledge base. :wave:
Yes, OpenAI has indeed developed vision models. Specifically, the GPT-4-Vision-Preview model is available and can be accessed via the API. However, it's important to note that this model is not available via the playground. You should be able to use this model by making the appropriate API calls.
For more detailed information on how to use the vision models, you can refer to the OpenAI documentation (https://platform.openai.com/docs/guides/vision).
Also, OpenAI released GPT-4 Turbo with Vision in general availability in the API on Apr 9th, 2024. You can find more about this in the OpenAI Changelog (https://platform.openai.com/docs/changelog/released-gpt-4-turbo-with-vision).
Please ensure that you're using the correct API endpoint for your needs. For instance, one user found that they were able to retrieve and use the GPT-4-Vision-Preview model via API after they switched from using the completion API endpoint to a chat endpoint.
Sources:
- Forum Post: Access to GPT4 vision API > Official Answer (https://community.openai.com/t/624669)
- Forum Post: How to get access to gpt-4-vision-preview model? > Official Answer (https://community.openai.com/t/642861)
- Docs > Changelog > Apr 9th, 2024 (https://platform.openai.com/docs/changelog)