llmpress reduces the token cost of sending source code to an LLM by detecting repeated patterns, replacing them with short aliases, and transparently restoring the originals in the model's response.
By chatting or signing in you agree to the Terms and chat-message logging (revocable in History).