New APIs for ChatGPT and Whisper open up growth and business opportunities amid the biggest buzz in tech.
Startup OpenAI has launched an API for developers and enterprises to embed their ChatGPT and Whisper models into apps, websites, and their products and services.
The innovative new firm's joint statement released this week, including comments made by Greg Brockman, president and co-founder of OpenAI, confirmed that it should provide access to its advanced languages and speech-to-text capabilities. The statement said: “Through a series of system-wide optimizations, we have achieved a 90% reduction in ChatGPT costs since December; we are now passing these savings on to API users.
“Now developers can use our open source Whisper large-v2 model in the API with much faster and more cost-effective results. Users of the ChatGPT API can look forward to continuous model improvement and the choice of dedicated capacity for greater control over models.”
The move to API discovery is huge and points to OpenAI's strategy to get their money back. Since its release in December, ChatGPT is estimated to have gained over 100 million users. However, many wondered where it would bring the return on investment.
The OpenAI proposal means that firms can develop improvements and integrations with a choice of models and ways to get the service. But with OpenAI projected to bring back just $200 million in 2023, the company is forced to capitalize on the hype surrounding its technology.
The statement added: “We believe that AI can provide incredible opportunities and economic opportunities for everyone. The best way to achieve this is to let everyone build with it. We hope that the changes we announced today will lead to numerous applications that everyone can use.”
Introduction to ChatGPT Plus
Driven by unprecedented global interest, and looking to recoup the investment made by Microsoft, which pumped another $10 billion into the technology in January, it launched the ChatGPT Plus subscription pilot plan. The new tier was launched at $20/month, with the free ChatGPT tier remaining.
In a blog post about the release of ChatGPT Plus, the firm stated, “We love our free users and will continue to offer free access to ChatGPT. By offering this subscription price, we can help provide free access to as many people as possible,” the company continued. "We plan to refine and expand this offering based on your feedback and needs."
ChatGPT and Whisper API pricing
The ChatGPT family of models released this week is still the 3.5 Turbo model, but will be ten times cheaper at $0.002 per 1,000 tokens. The startup said that this current version of the "0301" model will be supported until June 1st.
The Whisper Transcription Model API will cost developers $0.006 per minute, provide "robust" transcription in multiple languages, and provide English translation.
Early global users of ChatGPT and Whisper
With the launch of the API, OpenAI has released a list of the first users of the ChatGPT and Whisper models with impressive implementation statistics.
The creators of Snapchat Snap introduced My AI to their 750 million monthly Snapchat users for a start this week.
Global learning platform Quizlet will introduce Q-Chat, a fully adaptive AI tutor designed to attract a 60 million student user base, after three years of GPT-3 rollout across multiple use cases.
Shopping app Instacart will use ChatGPT with its own artificial intelligence program for its more than 75,000 retail partner stores to use product data to answer customer questions about recipes.
Used by 100 million shoppers, the Shopify Shop consumer app will be deployed to help shoppers with shopping recommendations by scanning millions of products in a database.
Language learning app Speak is the fastest growing English learning app in South Korea that uses Whisper for open speaking practice.
Dedicated Developer Instances
The firm also offers "dedicated instances" for those seeking system control and performance. Azure Provisioned Instances APIs mean that developers are not charged per request, but rather per allocation of compute infrastructure over a period reserved for service requests. Developers can enable a dedicated instance, and OpenAI has suggested that this is cost-effective for those who use more than 450 million tokens daily.
Developers will be able to make applications perform a number of tasks, such as: composing an email or other text, assembling Python code, answering questions about various documents, creating a dialog agent, creating interface in natural language, etc. d. tutor in various subjects, translates language and models characters such as video games.
It looks like OpenAI is betting on creating a platform on which businesses can build their business. To that end, the startup focused on feedback and offered incentives for rewards. It subsequently made some changes to its terms, such as making it easier to use the data and clarifying deposit and withdrawal information for users. It also automates the pre-launch verification process.
AI in Microsoft 365 and the Teams platform
Earlier we looked at how the potential of AI models means that developers can use the APIs in Microsoft Teams to build a chatbot. The same article also shows how Teams is already using AI to prepare meeting benefits such as debriefing, scheduling activities, and creating chapters.
After that, Microsoft announced an integration for Teams Premium and added AI to meetings with features like Intelligent Recap, Live Translation, and Templates Meeting.
Microsoft has been publicly supporting AI and driving product integration for some time, from the introduction of the DALL-E 2 model across much of the Microsoft 365 ecosystem when DALL-E 2 was introduced across much of the Microsoft 365 ecosystem last fall, until recently time. introduces its Bing search engine with an extension based on ChatGPT. The current strategy has come to a point where Microsoft appears poised to take over Google's dominance of the search market.
Satya Nadella, Chairman and CEO of Microsoft, said: “AI will fundamentally change every category of software, starting with the biggest category, search. We are launching Bing and Edge on a co-pilot and AI chat to help people get more out of search and the web.”
Bing vs Bard
According to Statista, Bing currently holds 8.85% of the global search market, while Google's market share has dropped to 84.69% from over 90% in 2022.
The Bing release seems like part of a tennis match to see who integrates the search chatbot faster. Last month, Google also released an artificial intelligence model for Bard storytelling. The model is based on LaMDA (Language Model for Conversational Applications) and has been seen as a reflection of Microsoft ChatGPT since December. However, Google has been developing AI for more than two decades and has about twenty sets of tools for developers who want to use open source and API services.
Comments