Rumor has it that ChatGPT update will have video and audio features

ChatGPT’s latest large language model (LLM) GPT-4.5 update was leaked before release. Among the details disclosed; the robot must include images, video, audio and also 3D capabilities. OpenAI, the developer of artificial intelligence (AI), has not yet confirmed the information.

An excerpt from the leaked draft was posted on X, formerly Twitter, by user Daniel Nguyen. According to him, the update will enable ChatGPT to have an interconnected understanding between different forms of interaction.

Did OpenAI “leak” GPT-4.5?

The state-of-the-art model delivers complex reasoning and cross-modal understanding, as well as multimodal capabilities across language, audio, image, video, and 3D.

New models:

• GPT-4.5

• GPT-4.5-64k

• GPT-4.5 audio and speech pic.twitter.com/hRyAbJXK40

— Daniel Nguyen (@daniel_nguyenx) December 14, 2023

GPT-4.5 is OpenAI’s most powerful model. The cost of the update will be $0.06 per thousand login tokens, or approximately R$0.29. The cost of departures will still be US$ 0.18 or R$ 0.88 per thousand. The leak also contains other values. are all higher than the GPT-4 version used in ChatGPT Plus and Microsoft Copilot:

• GPT-4.5 64K – $0.12 per thousand coins in and $0.36 per thousand coins out
• GPT-4.5 Voice and Speech – $0.012 per minute and $0.024 per minute throughput

Since OpenAI has not commented on the leak, it is difficult to confirm the accuracy of the information. If the document is correct The new version of the language model will be released in the next few daysAs with GPT-4 Turbo.

It’s also worth remembering that it came after Google announced Gemini. Big tech’s AI is capable of understanding various types of combined input; This was the previous weakness of the startup’s chatbot. The model is now available in the Bard chatbot.