OpenAI releases its GPT-4 API in general availability, giving all paying developers access and planning to give new developers access by the end of July 2023
OpenAI today announced the general availability of GPT-4, its latest text-generating model, through its paid API.
Context & Ripple Effects
OpenAI had already turned GPT-3 from a waitlisted product into a broadly available API in supported countries through its GPT-3 API general release. GPT-4 initially followed a more controlled path, launching with API access behind a developer waitlist.
This move shifts GPT-4 from constrained rollout to a standard paid developer offering. It matters because availability—not merely a model announcement—determines whether developers can build and operate products around it.
First-order effects
- Paying developers can integrate GPT-4 without waiting for individual API approval, while prospective new developers are slated for access by the end of July.
- OpenAI moves GPT-4 closer to a production platform product, making paid API access the immediate route for developers seeking the model.
Second-order effects
- Application builders can plan GPT-4-based features around a more predictable access path, reducing the advantage held by early API recipients.
- The release raises the competitive bar for AI API providers: access policies and developer onboarding become part of the product comparison, alongside model capabilities.
Third-order effects
- If repeated across model generations, frontier-model launches may increasingly follow a two-stage pattern: restricted initial access followed by paid API commercialization.
- That pattern would make API providers more central gatekeepers of advanced-model distribution, with access terms and rollout timing shaping which developers can deploy new capabilities first.
The trend: This is one data point in the platformization of generative AI, where leading models move from limited previews into paid, developer-facing infrastructure.