OpenAI Has Introduced ChatGPT As Well As Whisper APIs To Developers
OpenAI has made APIs for developers available for its well-known ChatGPT as well as Whisper AI models. This allows app developers to enhance their apps’ capabilities with OpenAI technology.
The API (application programming interface) is an established set of protocols that allow different computer programs to communicate with one another. The cost for the integration of the models in their applications is based on the amount of use.
OpenAI Launches API for Whisper
OpenAI announced that it has launched an API that supports Whisper the model for speech-to-text that the AI research firm opened-sourced in September 2022. The API is supported by ChatGPT that generates consistent text in a variety of styles. Whisper helps translate spoken words into text.
API changes for developers
The company also announced that developers have access to the latest capabilities for speech-to-text and language via it’s ChatGPT as well as Whisper model available on the API.
Additionally, it stated the fact that there was an improvement of 90% in cost from December to date, meaning that users can now utilize Whisper, an open source Whisper large-v2 model within the API with significantly faster and cost-effective outcomes. It also includes additional features including:
Continuous model improvements
Select dedicated capacity for greater control
The API terms of service have been refined for developers
Early adopters of ChatGPT and Whisper APIs
Snap Snapchat’s creators Snapchat revealed My AI, a chatbot for Snapchat+, an experimental feature powered by the ChatGPT API earlier this week. With this user-friendly chatbot that can be customized, Snapchatters can get personalized recommendations , and much more.
Quizlet is leveraging openAI’s GPT-3 to power a variety of use scenarios, including test preparation and vocabulary learning. With the release of the ChatGPT API Quizlet introduces Q-Chat as an adaptive AI tutoring system that communicates with students by providing relevant learning materials and presented in a fun chat-based experience.
Instacart 2.0: Now, with ChatGPT, Instacart is taking its application to the next step, by introducing ChatGPT as well as Instacart’s own AI that provides its “inspirational, shoppable answers” to questions from customers about food items. With access to more than 75,000 locations of retail partners that customers can access everything they require to fulfill their ever-changing shopping needs.
Shopify Shopping powered through the ChatGPT API, it is the basis for Shop’s brand new shopping assistant, which makes individual recommendations based on customers’ requests for search results.
Speak has been developed using AI to create a app for learning languages that provides the most effective path towards fluency in spoken language. The app is the most popular and fastest growing English application on the market in South Korea, and is currently making use of its Whisper API to develop an AI speaking product.
OpenAI has introduced their ChatGPT models, including the gpt-3.5-turbo one, that is the exact model that is used to create ChatGPT. The price is $0.002 per token (1k) and is 10x cheaper than the GPT-3.5 models, which makes it the preferred model for a variety of non-chat instances. Early testers have found the transition from text-davinci 003 to gpt-3.5-turbo to be a breeze, only making minor adjustments along the way to get the best output.
ChatGPT models consume a series of metadata and messages, that are displayed to the model as the sequence of “tokens” using a new format known as Chat Markup Language (“ChatML”).
“In dedicated instances, users can have complete control over the workload of the instance and can choose to activate features like longer limit on the context of the instance, and the option to save to the snapshot of their model. It can be beneficial economically for users who use more than 450 million tokens per day, since it could dramatically cut expenses compared to shared infrastructure.”
Text-to-speech made simple, Get quick, instant accessibility to the Whisper API’s large-v2 model priced at $0.006/minute. Get enhanced functionality and support for range types of files (m4a, MP3 (mp4, mpeg) wav, mpga webm).
OpenAI stated that it has been paying attention to developers for the last 6 months gathering feedback and making changes that will help provide the best service possible:
API data submission is not required unless the your organization has opted in.
The default 30-day data retention policy and options for more stringent retention.
Eliminated pre-launch review.
Improved developer documentation.
The simplified Terms of Service, as well as Usage Policies, which include data ownership provisions.
Starting today, you can use the ChatGPT as well as the Whisper APIs. For more details.