Three ways to work with llamamon
Pick the mode that fits your workflow — all three share the same config backbone and hardware-detection layer.
- Full model list with VRAM estimates
- 5-tab config visible before launch
- Live hardware stats while server runs
- Best when actively exploring local setups
- TOML model definitions are versioned
- Reproducible across machines and teams
- Great for hardware-specific tuning handoffs
- Fits CI environments and benchmarking
What works where
llamamon vs the alternatives
Ollama runs models. nvidia-smi watches heat. Datadog monitors servers. llamamon does all three — and then translates the telemetry into financial and compliance intelligence.
Pro = Professional edition · Ent = Enterprise edition (contact sales) · Custom = bespoke internal tooling estimate
Local vs cloud — in the TUI
Session cost, cloud price comparison, and hardware amortization — measured from real runs, displayed live.
Deeper comparisons
Professional capabilities are available now. Enterprise comparison layers expand through demand-driven design partner rollouts.
Compare a local run against cloud API pricing, electricity rates, and hardware amortization — not just throughput.
Project what a measured single-node benchmark implies for a larger workstation or cluster purchase.
Understand whether a model is commercially safe, attribution-bound, or pending legal review before production.