How Cloud / SaaS and on-premises modes work, why assumptions stay conservative, and how to embed the standalone tool. Per-seat plans are modeled here; Enterprise commercial terms depend on your deployment, scale, and support requirements.
Same build as the standalone file.
Six inputs: team size (users) and five monthly spend lines (LLM API, agent/orchestration, ML/evaluation, chat/productivity, other). The five spend lines are summed and annualized (×12) for current annual spend.
Compare to VDF AI Starter ($20/user/mo) or Professional ($50/user/mo). Enterprise pricing depends on factors such as deployment model (cloud vs on-prem), scale, integrations, and support—so it is agreed with your team rather than a single list price here.
The comparison highlights multi-vendor stacks (e.g. cloud APIs + orchestration + ML platform + copilots) versus consolidated per-seat platform pricing.
Model-provider token spend may still apply; routing optimizations are modeled separately from this “replace the SaaS stack” view.
Six annual inputs plus a routing-efficiency control for energy: energy, DevOps & ML engineering, model evaluation, orchestration licences, downtime/incident cost, and other annual spend (no modeled reduction on that line).
Reduction rates on modeled categories: energy (your 30–65% routing band), DevOps/ML 45%, evaluation 80%, orchestration 60%, downtime 50%.
The tool shows operational savings vs status quo after those levers. The VDF Enterprise platform fee depends on your environment and contract scope; net economics including platform are finalized with your team.
Each row ties a mechanism to a percentage so finance can audit the logic.