Subconscious Raises $5.1 Million to Build the Inference Platform for Long-Running Agents

Subconscious has raised $5.1 million to launch an inference platform specifically designed for long-running AI agents. The platform is based on MIT research and uses dynamic context compression to make agents run faster, for longer, and at a significantly lower cost. This technology can compress up to 80% of an agent's context, extending the effective context window of models and improving performance on key benchmarks. The system allows engineering teams to use powerful open models more efficiently, drastically reducing AI operational costs while improving task accuracy.