GPT-4, a highly capable model, is being introduced to API users today, with a live demo scheduled for March 14th at 1 pm PDT. The model boasts broader general knowledge and advanced reasoning capabilities, allowing it to solve complex problems accurately. More information can be found on the GPT-4 overview page and blog post.
API waitlist access is available for GPT-4, with rate-limited access using the same ChatCompletions API as gpt-3.5-turbo. Developers contributing model evaluations to OpenAI Evals can get priority access. ChatGPT Plus subscribers will have access to GPT-4 on chat.openai.com with a dynamic usage cap.
API pricing for gpt-4 with an 8K context window is $0.03 per 1K prompt tokens and $0.06 per 1K completion tokens. For gpt-4-32k with a 32K context window, pricing is $0.06 per 1K prompt tokens and $0.12 per 1K completion tokens.