← back

# RLM-TS

TypeScript implementation of Recursive Language Models (RLMs) — an inference paradigm where models treat their own prompts as objects in an external environment, writing code that invokes LLMs to recursively decompose and process long contexts.

Built on the Vercel AI SDK, works with any compatible provider (OpenAI, Anthropic, Google). Supports local and Docker sandboxed execution environments. Includes a debugger for visualizing execution trajectories.