Meta Releases Llama 3 to Challenge Top Proprietary AI Models
Meta introduces Llama 3, a new open-source AI model designed to compete directly with leading paid systems from OpenAI and Google. The company also integrates this upgraded technology across its major social media platforms.
Meta releases Llama 3, an updated open-source artificial intelligence model designed to compete directly with top proprietary systems like GPT-4 and Claude 3. The company currently launches two smaller versions of the model across major cloud platforms, positioning them as the best open-source models in their class.
The new models feature improved reasoning and coding capabilities, along with plans for future multimodal and multilingual functions. Meta also aims to provide larger context windows to reduce AI hallucinations and allow the system to process more data at once.
A much larger 400-billion-parameter version remains in training and shows potential to match the best closed-source models on the market. Alongside this release, Meta integrates an upgraded Meta AI assistant directly into the search bars of WhatsApp, Instagram, Facebook, and Messenger.