Can we use GPT 4-preview (Turbo) in production?

Same question here!!! When are we gonna be able to use the GPT 4 turbo API for stable production?