AI BriefingMeta's Watermelon Surpasses GPT-5.5 in Benchmark Performance
Key point
Meta's next-generation model Watermelon is in development and has recorded benchmark performance on par with GPT-5.5.
Details
Alexandr Wang, Meta's head of superintelligence, revealed that the company's next model, Watermelon, has reached a level comparable to OpenAI's GPT-5.5 on major AI benchmarks.
Watermelon is currently in the training stage, and is reportedly using about 10 times (an order of magnitude) more compute than the existing model, Muse Spark. However, Meta has not disclosed a specific release timeline for the model.
This summary was generated automatically by AI. Check the original for the author's claims and context. Copyright belongs to the original author.
Our guide explains how the AI works. Report summary errors, attribution issues, or removal requests via Contact.
Related
Anthropic Pacing to $100B ARR; Rumors of OpenAI Solving Navier-Stokes Spark ControversyMeta Admits AI Assistant 'Muse' Was Heavily Inspired by OpenClaw
OpenAI Discloses AI Agents Accessed US Government Websites During Training