Today Meta introduced its latest large language model, Llama 3, a real time text to image generative AI and an image generator that “updates pictures in real time while users type prompts.” The AI products are set to eventually challenge the current generative AI market leader OpenAI.
The models will be integrated into its virtual assistant Meta AI, which the company is pitching as the most sophisticated of its free-to-use peers, citing performance comparisons on subjects like reasoning, coding and creative writing against offerings from rivals including Alphabet’s (GOOGL.O), opens new tab Google and French startup Mistral AI.
Although Llama 3 currently outputs only text, the model will eventually become multimodal and generate both text and images:
The goal eventually is to help take things off your plate, just help make your life easier, whether it’s interacting with businesses, whether it’s writing something, whether it’s planning a trip.
Compared to previous model Llama 2, Llama 3 uses “high quality data” and seven times more data for its textual output. Llama 2 reportedly was often unable to understand basic context.
Llama 3 is a dramatic generative AI improvement as Mark Zuckerberg recently stated at the its generative AI products announcement:
Zuckerberg said the two smaller versions of Llama 3 rolling out now, with 8 billion parameters and 70 billion parameters, scored favorably against other free models on performance benchmarks commonly used to assess model quality. The biggest version of Llama 3 is still being trained, with 400 billion parameters, he said.
