Feature comparison
How credits work
Credits are consumed per analysis based on type and complexity. Cache hits cost 1 credit.
Questions
+What are the 7 engines?
Analysis, Execution, Self-Healing, Visual Intelligence (V-JEPA 2), Evidence Chain, Training (QLoRA fine-tuning), and Inference (per-tenant model serving). Each engine feeds the next.
+How does per-tenant training work?
On Scale, every analysis harvests training data from your codebase. This data fine-tunes a QLoRA adapter on your patterns. Your custom model is then served for future test generation.
+Can I cancel anytime?
Yes. Cancel anytime and keep access until the end of your billing period. No long-term contracts on Cloud plans.
+Cloud vs Self-Hosted?
Cloud runs on our infrastructure with zero setup. Self-hosted deploys to your K8s cluster — including restricted-network deployments with Ollama for offline LLM inference (air-gapped available by enterprise agreement).
Need a custom deployment?
Self-hosted, restricted-network, or custom cloud — we deploy QuantumVerifi where your code lives (air-gapped available by enterprise agreement).
Contact Enterprise SalesAll prices in GBP, exclude VAT where applicable (UK: 20%). QuantumVerifi Ltd · VAT: GB 507 1047 26