Meta Trains Llama 5 Codenamed Watermelon
Meta is reportedly prepping its next massive frontier model to compete directly with OpenAI. Codenamed Watermelon and believed to be Llama 5, the model is currently in training using roughly ten times the compute power of April's Muse Spark release.
Meta is significantly scaling its compute resources to keep pace with OpenAI's frontier developments.
- –A 10x compute scaling factor suggests Llama 5 will be dramatically larger and more capable than previous iterations.
- –Codenaming the project Watermelon shows Meta is actively developing distinct branches of its LLM family.
- –The leak indicates that Meta has no intention of slowing down its high-compute training runs despite rising infrastructure costs.
DISCOVERED
1d ago
2026-07-06
PUBLISHED
1d ago
2026-07-06
RELEVANCE
AUTHOR
LuminaXspace