Meta is reportedly developing a new artificial intelligence model codenamed “Avocado”, which will replace the existing Llama model. Some rumors also say that the model will be exclusive to Meta, and will no longer be offered in open source form like the previous Llama model.
In addition to the model for text, Meta is now reportedly developing a new model for images and videos, and will be introduced in the first half of 2026. With the new model, it is expected to offer better image and video generation – likely in line with what Google and OpenAI offer.
With these various developments, it is expected that 2026 will be the year in which Meta focuses entirely on the artificial intelligence arena. The Meta AI that was introduced is also still not very successful in stealing the attention of many, and does not provide good responses to various questions. Let's look forward to how the various developments of this new model will lead to improvements in Meta AI.