Vector Completions

The core ObjectiveAI primitive. Takes a prompt and multiple responses, then uses an ensemble of LLMs to vote and rank them.

Input

tex
tex

Output

📊

Execute to see ranked results

How it works

  • Each LLM in the ensemble votes on which response best answers the prompt
  • Votes are weighted by the profile to compute final scores
  • Scores sum to 1 - higher score means stronger preference
  • You need an Ensemble ID - browse ensembles at /ensembles