Delivery
Choose your personas and precision
Select the FuZeLLM personas you need and pick a numeric precision profile. We’ll compute directional estimates for FuZeCLOUD and FuZeBOX, then you’ll choose delivery.
What are you buying?
Choose the RRLM persona stack, or a simpler hosted LLM platform with no RRLM/personas.
Hosted LLMs
If you don’t want RRLM, we’ll host the LLM(s) you specify and operate the platform around them. Pricing excludes persona add-ons and scales with the number of hosted models.
Selected
0 models selected
Pick one or more models below. This list is derived from the union of models in our product build configs.
Choose model(s)
Loading curated model list…
Personas
Toggle the personas you want at launch. You can always adjust later during implementation.
None selected yet
Precision profile
Higher precision buys more numeric headroom. Quantized runs cheaper and faster.
Default: balanced
Performance target
Plans are priced for a 50 tok/s guarantee. Each additional +50 tok/s doubles the price.
50 tok/s (1×)
If you’re unsure, start with 50 tok/s and scale up later.
Directional estimate
These numbers are directional only. Final commercial terms are confirmed in your quote and paperwork.
Delivery: Not chosen yet
Build selection: —
Personas: None selected yet
Precision: Balanced (FP16)
Performance: 50 tok/s (1×)
Model band: S · ≤10B
FuZeCLOUD
$0/mo
Approximate monthly for a managed FuZeCLOUD tenancy with this configuration.
FuZeBOX
$0/mo
Deposit at signing: $0
Select at least one persona to proceed to delivery.