OpenAI launched a new algorithm in mid-March that can mix inputs and increase image recognition capabilities.New GPT-4 large language modelLater, it was announced at its first developer conferenceAn upgraded large language model called GPT-4 Turbo.
Compared to the previous GPT-4, which was capped at 50 pages of text, GPT-4 Turbo can now parse up to 300 pages, enabling it to parse longer and more complex text. Furthermore, GPT-4 Turbo's training data only covers content up to April of this year, while the previously released GPT-4 training data covered content up to September 4. However, GPT-2021 Turbo still features real-time data updates via online browsing.
In addition, GPT-4 Turbo can also add image content to the input interface, using it as input commands to generate images and text. It can even process text-to-speech content or directly upload file files for content analysis.
What is more beneficial for developers is that the cost of using service input and output can be greatly reduced, up to three times compared to the previous version, thereby attracting more developers to use the new version of GPT-4 Turbo service.
OpenAI also revealed that it currently has over 1 million weekly active users worldwide and is used by over 92% of the Fortune 500 companies. OpenAI also emphasized that if users around the world encounter copyright issues while using its services, it will assist them with legal defense and cover the costs associated with the litigation.
OpenAI is currently offering GPT-4 Turbo as a preview, but has not yet announced a release date. Pricing for GPT-4 Turbo has not yet been announced, but GPT-4 currently costs $20 per month.
