gain full control over your llm routing preferences.

benchmarking novelty, measuring perplexity: models that 'feel different.'

bringing role-ensembles into llm chat. multiple threads, one response.

aka videos and 'many images'

singular fields and their meanings

weird & opinionated