Free Scaling

$0 test-time scaling infrastructure using NVIDIA NIM free tier.

Three patterns, one API key:

from free_scaling import scale, generate, health

# Classify — vote on labels
result = scale("Is this safe?", context=code, k=3,
               answer_patterns=["SAFE", "VULNERABLE"])

# Generate — best-of-k with cross-evaluation
result = generate("Summarize this paper.", context=paper, k=3)