Memory
capacity—
- Weights
- [0.0GB]
- KV/token
- [0.0KB]
- Concurrent (req · max)
- 0 · [0]
- Context used
- —
- Free VRAM
- —
- Activation overhead
- [0.0GB]
Calculator
Pick a model and hardware, then move sliders. Numbers re-compute below.
Model
Hardware
GPUs
Avg input tokens
Avg output tokens
Concurrent users
Max context for memorycaps KV-budget
Throughput
bandwidth[0]
Prefill tok/s
[0.0]
Decode/user tok/s
[0]
Aggregate tok/s
- MFU prefill
- [0%]
- BW efficiency
- [0%]
- Effective batch
- [0]
Latency
time[0]ms
TTFT
[0.0]ms
ITL
—
End-to-end
- Effective batch
- [0]
Cost
economics[$0.00]
$/M input
[$0.00]
$/M output
[$0.0000]
$/request
- GPU $/hr (total)
- [$0.00]
- Monthly (730 hr)
- [$0]