OpenAI, the maker of the popular AI chatbot ChatGPT, has announced the release of its latest ChatGPT and Whisper models for developers to leverage the “cutting-edge language and speech-to-text capabilities.” The company has made significant system-wide optimizations resulting in a 90% cost reduction for ChatGPT since December. These cost savings are being passed on to API users, enabling them to use these open-sourced APIs for faster and more cost-effective results.
“We believe that AI can provide incredible opportunities and economic empowerment to everyone, and the best way to achieve that is to allow everyone to build with it. We hope that the changes we announced today will lead to numerous applications from which everyone can benefit,” OpenAI writes in its blog post.
ChatGPT API users can expect continuous model improvements and the option to choose dedicated capacity for deeper control over the models. Meanwhile, those using the Whisper API can expect faster and more cost-effective results. The company has released the new gpt-3.5-turbo model, which is the same model used in the ChatGPT product. It has been priced at $0.002 per 1k tokens, which according to the company, is 10 times cheaper than its existing GPT-3.5 models. OpenAI also claims that it is the best model for many non-chat use cases.
The company has listed the names of some of the early adopters of these APIs. Snapchat, the popular social media platform known for its disappearing messages, announced My AI this week. It is a ChatGPT API-based chatbot that can give recommendations and can even write a haiku for friends in seconds. Quizlet, a global learning platform, has also leveraged the GPT-3 technology for Q-Chat, a fully-adaptive AI tutor that engages students with adaptive questions based on relevant study materials, delivered through a fun chat experience. Instagram and Shopify are also using ChatGPT API on their platforms.
Additionally, Speak, an AI-powered language learning app, is using the Whisper API to power a new AI speaking companion product. First announced in September, Whisper is an automatic speech recognition (ASR) system that has been trained on 680,000 hours of multilingual and multitask supervised data collected and available on the web. It is basically a speech recognition model trained on a large and diverse dataset that can perform speech translation and audio transcription. The Whisper API is priced at $0.006 per minute by the company.
OpenAI has also made changes to its API services based on the feedback it received from its API customers. These changes include a default 30-day data retention policy for API users, removing pre-launch review, improving developer documentation, and simplifying its terms of service and usage policies.
ALSO READ l ChatGPT effect: Facebook, Snapchat, Twitter are joining the AI wars
