Meta introduces Llama-3 and Vision-X AI models for text, code, and image processing, challenging OpenAI and Anthropic with accessible hardware deployment.

Meta introduced two new artificial‑intelligence models on Monday, saying they are designed to narrow the gap with rivals such as OpenAI and Anthropic.

The models, dubbed Llama‑3 and Vision‑X, support text generation, code assistance and image‑to‑text conversion. Meta said the systems can run on a broader range of hardware, allowing developers to deploy them on‑premises or in the cloud.

"We are delivering next‑generation models that prioritize safety and accessibility," a Meta spokesperson said in a statement. "Our goal is to give creators powerful tools while protecting users from misuse."

Industry analysts noted that Meta's announcement marks the company's most aggressive push into generative AI since its 2023 partnership with OpenAI. "Meta's move shows it is accelerating its AI roadmap and seeking to compete on both performance and cost," said Jordan Patel, senior analyst at TechInsights.

Meta plans to integrate the models into its suite of products, including the Workplace collaboration platform and the upcoming AI‑enhanced version of its advertising engine. The company also announced a free tier for researchers and small businesses, aiming to broaden adoption.

OpenAI and Anthropic have recently expanded their own model offerings, intensifying competition across the sector. The race has prompted regulators to scrutinize AI safety practices, a concern Meta addressed by highlighting built‑in content filters and transparency tools.

Developers can access the models via Meta's AI Hub starting next week. Pricing details will be released later this month.

While Meta's models are not yet benchmarked against OpenAI's GPT‑4 or Anthropic's Claude, early tests suggest comparable fluency in language tasks and improved image understanding. The company said it will publish performance metrics in a forthcoming technical paper.