OpenAI modifies the costs of the real bees for vocal chatbots | OpenAI Playground | OpenAI Chat | ChatGPT app | Turtles AI

OpenAI has revealed innovations in its real bees that promise significant savings for vocal chatbot developers. With the new caching feature of audio and text inputs, the cost of long conversations could drastically decrease. The move is an attempt to make the use of AI in the sector more accessible.

Key points:

• Introduction of the cache storage for the reality API of Openai.
• Possible reduction in costs for conversations up to 80%long.
• API is already used by companies like Healthify, Speak and Twilio.
• The current input and output rates are considered high by some developers.

In an event dedicated to developers in London, OpenAI presented significant updates on its reality bees, making known new methods to optimize costs, particularly for vocal applications. This API, designed to facilitate the creation of virtual assistants and chatbots, has received mixed feedback from developers due to the high rates associated with input and output processing. The new functionality incoming include an automatic caching system for audio and text inputs, which allows you to avoid the reworking of the data already processed. This mechanism translates into a significant decrease in costs for prolonged conversations, which could undergo a reduction of up to 80%. Until now, every time a user interacted with a bot powered by this API, the system had to rework all the input and output tokens of previous interactions, making costs rise especially in multi-platern conversations. According to Kwindla Kramer, Daily co-founder, the reworking of all token makes the use of the API particularly expensive. Shaun Ralston, a member of the OpenAI API support team, clarified that thanks to the new caching functionality, it will be possible to recover the previous input results, applying reduced rates, thus avoiding the need to develop every part of the conversation again. The new rates, which will drop from 5 to 2.5 dollars per million textual tokens and 100 to 20 dollars per million audio token, represent an important step towards economic sustainability for developers. This innovation aims to make the Ai more accessible and convenient for a variety of uses in the sector, placing OpenAI in a competitive position in the rapidly evolution market of services AI.

An important progress that could transform the way the developers interact with vocal technologies.