ChainCatcher news, recently OpenAI CEO Sam Altman revealed on social media that the Alpha version of GPT-4o's Voice Mode will be available to ChatGPT Plus subscribers this week, to help users engage in seamless conversations without delay.According to the current public information, the ChatGPT Voice Mode has an average latency of 2.8 seconds (GPT-3.5) and 5.4 seconds (GPT-4), making the user experience less intuitive, while GPT-4o will significantly reduce the latency time, enabling seamless conversations.

ChainCatcher news, according to CNBC, the artificial intelligence startup OpenAI has launched a new AI model "GPT-4o mini," claiming that the new version is "the most powerful and cost-effective small model available today," with plans to integrate images, videos, and audio later.It is reported that the "o" in GPT-4o stands for omni, and GPT-4o has improved audio, video, and text capabilities, enabling it to process 50 different languages at higher speed and quality.

ChainCatcher news, Google held its annual developer conference "I/O 2024" early this morning, announcing a brand new AI agent product Project Astra, which can achieve cross-text, audio, and video multimodal real-time reasoning, similar to OpenAI's released GPT4o.It also launched an update for Gemini 1.5 Pro, whose AI model will soon be able to handle more data, along with a new smaller Gemini 1.5 Flash AI model.Additionally, Google released its latest model "Veo" for generating high-definition videos, as well as the highest quality text-to-image model Imagen 3, which promises to deliver realistic images and "less distracting visual artifacts than our previous models."

ChainCatcher news, OpenAI has released its latest flagship model GPT-4o, which can perform real-time reasoning on audio, visual, and text inputs, focusing on the concept of a personalized, supernatural, ultra-low latency voice interaction assistant.According to information from the official OpenAI website and the official account on the X platform, the "o" in GPT4o stands for Omni (comprehensive), marking a step towards more natural human-computer interaction—accepting any combination of text, audio, and images as input and supporting the generation of any combination of text, audio, and image outputs. It can respond to audio input in 232 milliseconds, averaging 320 milliseconds, which is similar to human response times in conversations. In terms of English and code, it performs comparably to GPT-4 Turbo, with significant improvements in non-English language text, while the API is faster and 50% cheaper. Compared to existing models, GPT-4o excels particularly in visual and audio understanding. Text and image inputs will be launched today in the API and ChatGPT, while voice and video inputs will be rolled out in the coming weeks.