OpenAI released four new tools on Tuesday that would make it easier for developers to create applications based on its artificial intelligence technology, as the ChatGPT manufacturer competes with IT behemoths to stay ahead in the generative AI race.
First, it launched a public beta of the Realtime API, which would enable developers to create low-latency, multimodal experiences for their apps.
This will allow developers to create applications comparable to ChatGPT’s Advanced Voice Mode, which features natural speech-to-speech chats with six preset voices.
In addition, as part of Tuesday’s deployment, the Sam Altman-owned AI business introduced a fine-tuning tool for models after training, allowing developers to improve model responses using images and text.
Third, OpenAI is now catching up to Google and Anthropic with the new Prompt Caching capability, which enables developers to cut costs and latency.
Developers can save 50% and speed up their processing times by using Prompt Caching.
Furthermore, developers do not need to do anything to activate Prompt Caching; it is automatically applied to the most recent versions of GPT-4o, GPT-4o small, o1-preview, and o1-mini, as well as fine-tuned versions of those models.
Finally, OpenAI unveiled a new package of tools called Model Distillation, which will let programmers use the outputs of larger frontier models to fine-tune smaller models.
Newsng understands that DevDay isn’t being webcast live like it was last year, but OpenAI still intends to upload videos to its YouTube page afterwards.
Demos, community spotlights, and breakout sessions are all part of the event’s schedule.
The CEO of the company did not give a keynote address this year, which is the biggest difference from the previous one.
The OpenAI product team was in charge of this year’s keynote.
We earlier reported that China, which leads the world in EV sales already, is now aiming to take the lead in the humanoid robot industry and may even surpass Tesla in the race to automate labour forces in factories.
