Welcome to the era of AI revolution, where OpenAI is paving the way with groundbreaking advancements and competitive pricing in the generative AI realm. In the face of escalating rivalry, the AI giant is fortifying its text-generating models and adopting a wallet-friendly approach for developers.
Unveiling the latest iterations of GPT-3.5-turbo and GPT-4, OpenAI brings to the fore a revolutionary tool - function calling. This remarkable feature enables developers to delineate programming functions to these cutting-edge models and watch in awe as they craft code to execute these functions seamlessly.
Imagine the possibilities! Function calling can empower developers to design interactive chatbots that answer queries through external tools, translate natural language into database inquiries, and extract structured data from text. OpenAI maintains, "Function calling allows developers to more reliably get structured data back from the model.”
Adding another feather to its cap, OpenAI is launching an enhanced variant of GPT-3.5-turbo equipped with an incredibly enlarged context window, a parameter measured in tokens or chunks of text. A large context window retains the model's memory of recent conversations, preventing off-topic detours that can compromise coherence.
This revamped GPT-3.5-turbo comes packed with a context window quadruple the size of the standard model (16,000 tokens), priced at twice the rate. However, the silver lining here is OpenAI's announcement of a 25% price reduction for the original GPT-3.5-turbo model, allowing developers to utilize it for around 700 pages per dollar.
OpenAI's generous price cuts also extend to text-embedding-ada-002, a model favored for determining text relatedness and employed for search and recommendation functions. Owing to heightened system efficiency, OpenAI now offers this model at a striking 75% price reduction.
Charting the course for future enhancements, OpenAI's focus remains on incremental updates over designing entirely new models from the ground up, as reaffirmed by CEO Sam Altman. This commitment reflects in the monumental strides OpenAI has taken since the release of GPT-4 in March, underlining the immense potential that the future holds for generative AI.
