What is Chat GPT all about
ChatGPT creator OpenAI announced the launch of GPT-4o, its newest flagship OPenAI's AI model. During a livestream on Monday (May 13, 2024) Company CTO Mira Murati described GPT-4o as "much faster" and "improves capabilities across text, vision, and audio" she further said- "this is incredibly important, because we're looking at the future of interaction between ourselves and machines." Also, the model will be accessible to all users.
According to OpenAI, the 'o' represents omni, GPT-4o marks a significant step towards more natural human-computer interactions. It can accept inputs in any combination of text, audio, and images, generating corresponding outputs in these formats. It can respond to audio inputs in as little as 232 milliseconds, with an average of 320 milliseconds response time, which is similar to human response time in a conversation.
It matches GPT-4 Turbo performance on text in English and code, with significant improvement on text in non-English languages, while also being much faster and 50% cheaper in the API. GPT-4o is especially better at vision and audio understanding compared to existing models.
GPT-4o greatly improves the experience in OpenAI's AI-powered chatbot, ChatGPT. GPT-4o can look at a picture of a menu in a different language and translate it. users can ask the GPT-4o-powered ChatGPT a question and interrupt ChatGPT while it's answering. The model delivers "real-time"responsiveness. OpenAI says, and can even pick up on tone of user's voice, in response generating voices in multiple emotions. GPT-4o is more multilingual as well, OpenAI claims, with enhanced performance in around 50 languages.
Features of GPT-4o
It builds upon the previous language models and with several enhancements. Here are some key points about GPT-4o:
Improved Intelligence and Capabilities
GPT-4o provides GPT-4-level intelligence but is much faster and more capable across text, voice, and vision. It excels at understanding and discussing images. Future improvements will enable more natural, real-time voice conversations and even real-time video interactions.
Language Support and Accessibility
GPT-4o supports more than 50 languages across sign-up, login, user settings, and more.
Free Access and Usage Limits
Features available to ChatGPT Free users include:
- GPT-4-level intelligence
- Responses from both the model and the web
- Data analysis and chart creation
- Photo discussions
- File uploads for summarization, writing, or analysis
- Use of GPTs and the GPT Store
- Building a more helpful experience with Memory
Difference between Chat GPT
Prior to GPT-4o What Chat GPT offering
You could use "Voice Mode" to talk to ChatGPT with latencies of 2.8 seconds for GPT-3.5 and 5.4 seconds for GPT-4 on average.
- Voice Mode has a pipeline of three separate models To achieve this:
- One simple model transcribes audio to text.
- GPT-3.5 or GPT-4 takes in text and outputs text.
- Third simple model converts that text back to audio.
With GPT-4o
Open AI trained a single new model end-to-end across text, vision, and audio, meaning that all inputs and outputs are processed by the same neural network. Because GPT-4o is first model combining all of these modalities, as per openAI they are still just scratching the surface of exploring what the model can do and its limitations.
Model evaluations
As measured on traditional benchmarks, GPT-4o achieves GPT-4 Turbo-level performance on text, reasoning, and coding intelligence, while setting new high watermarks on multilingual, audio, and vision capabilities.
Language tokenization
These 20 languages were chosen as representative of the new tokenizer's compression across different language families. shows outstanding Compression improvement for all 20 languages. for two languages see below.
English 1.1x fewer tokens (from 27 to 24) - Hello, my name is GPT-4o. I'm a new type of language model, it's nice to meet you!
Hindi 2.9x fewer tokens (from 90 to 31) - नमस्ते, मेरा नाम जीपीटी-4o है। मैं एक नए प्रकार का भाषा मॉडल हूँ। आपसे मिलकर अच्छा लगा!
Desktop App Integration
OpenAI has launched a new ChatGPT desktop app for macOS, designed to seamlessly integrate with users computer workflows. The ChatGPT app syncs your conversations, supports voice input, and brings our latest model improvements to your fingertips.
Model availability and pricing
GPT-4o is latest step in pushing the boundaries of deep learning, this time in the direction of practical usability. GPT-4o's capabilities will be rolled out iteratively. GPT-4o's text and image capabilities have been starting to roll in ChatGPT.
A new version of "Voice Mode" with GPT-4o Will be roll out in alpha within ChatGPT Plus in the coming weeks. At present, voice isn't a part of the GPT-4o API for all customers. OpenAI, citing the risk of misuse, says that it plans to initially launch support for GPT-4o's new audio capabilities to "a small group of trusted partners" in the coming weeks.
You can access from mobile app or from web.
To access ChatGpt-4o over web go to same Chat GPT URL
Click on Try it Now , as already discussed, you can switch free to plus subscription for extra benefit. or continue with free chat.
Note : "Voice Mode" with GPT-4o Will be roll out in alpha within ChatGPT Plus"
Conclusion
GPT-4o represents a significant step toward more natural human-computer interactions, accepting inputs in any combination of text, audio, and images, and generating corresponding outputs in these formats.
No comments:
Post a Comment