Recursive Language Models: The Future of Long-context LLMs

https://arxiv.org/abs/2512.24601 https://alexzhang13.github.io/blog/2025/rlm/ Recursive Language Models (RLMs) represent a significant advancement in large language model (LLM) inference strategies, introduced in a December 2025 arXiv paper by Alex L. Zhang, Tim Kraska, and Omar Khattab from MIT. The core innovation addresses the limitations of fixed context windows in LLMs, where performance degrades ("context rot") as input length increases beyond the model's native capacity, often around 128K-2M tokens for frontier models like GPT-5. Instead of forcing the entire prompt into the model, RLM…