OpenAI Opens GPT-4 API Access to All Existing Developers
OpenAI grants full API access to GPT-4 for existing developers with successful payment histories. The company plans to onboard new developers soon and introduce model fine-tuning capabilities later this year.
OpenAI officially grants general availability of GPT-4 through its API for all existing developers with a successful payment history. The company plans to welcome new developers by the end of this month and gradually increase access limits based on compute availability. This highly anticipated release follows millions of developer requests since the model's initial reveal in March.
The GPT-4 model processes both text and image inputs to generate text and code, outperforming its predecessor, GPT-3.5, across various professional and academic benchmarks. However, OpenAI currently restricts the image-understanding feature to a single partner, Be My Eyes, and does not specify a timeline for a wider rollout. Like other generative AI systems, GPT-4 still experiences hallucinations, reasoning errors, and difficulties with complex coding tasks.
OpenAI promises to introduce fine-tuning options for both GPT-4 and GPT-3.5 Turbo later this year, allowing developers to customize the models with their own data. This release arrives as competition in the generative AI space intensifies, with rivals like Anthropic recently expanding the context window of their Claude model to 100,000 tokens, surpassing GPT-4's 32,000-token limit.