Performance estimate

See time to first token, throughput, and memory estimates in seconds — refine as needed.

Model — Hugging Face ID

Checking...

Tested: , ... — type to autocomplete

GPU system

Based on your configuration — ISL 2048, OSL 128, FP16 KV cache, 32 concurrent users.

Select a model and GPU system above, then press Calculate to see results.

Want to change assumptions?

Every number above comes from these. Open a section to tune it — closed sections show their current values.