PRL3 · Adaptive Inference Suppression · Patent pending · USPTO #64/006,312

Benchmark Results

What PRL3 delivers

Fewer API calls. Lower costs. Same answers.

PRL3 is a drop-in proxy that sits between your application and your LLM provider. It intercepts incoming queries and resolves as many as possible without forwarding them to the LLM. Only queries that genuinely require the model are sent through.

The system learns continuously from live traffic. The more queries it processes, the higher the resolution rate becomes — without any manual configuration.

96.6%

API calls eliminated

94%

Answer accuracy

16×

LLM call reduction

~70%

Energy saved

96.6% vs 71.8%: The 96.6% figure reflects a warmed system operating on real repetitive chatbot traffic — the realistic production scenario. The 71.8% figure was measured during a cold-start benchmark of 830 fresh queries with no prior learning. Both numbers are real and documented. In production, as the system absorbs your traffic patterns, resolution rates converge toward the higher figure.

Accuracy benchmark

94% accuracy. 97% → 6% LLM dependency.

A 100-question dataset covering factual, definitional, computational, and multilingual queries was used to measure both answer quality and LLM call reduction across two evaluation passes.

In the first pass (R2), the system learned from scratch — 97 out of 100 queries required the LLM. In the second pass (R3), run after a 70% cache purge, only 6 queries required the LLM. Accuracy held at 94% across both passes.

100

Questions evaluated

94%

Accuracy R2 and R3

97 → 6

LLM calls R2 → R3

Methodology

Independently reproducible. Full install script public.

Parameter	Detail
Accuracy dataset	100 manually curated questions — fact, definition, calculation, multilingual
Throughput dataset	1,000 queries simulating real chatbot traffic with natural repetition
LLM backend	Groq API (cloud) · Local Qwen 2.5 1.5B (air-gapped mode)
Semantic verification	MiniLM 384d embeddings · similarity threshold 0.55 for cache admission
R2 → R3 protocol	Full learning pass → 70% cache deletion → validation pass on same dataset
Hardware	Intel i5-6600 · Quadro K2000 · Ubuntu 22.04 · kernel 5.15
Reproducibility	Full install script at icomnewtechnologies.com/prl3/

See for yourself

📹 Watch benchmark recording → Install PRL3 → Pitch deck → Contact →