local intelligence

can it run

Can the NVIDIA RTX A6000 run kokoro 82M?

FULL

2.3 GB needed of 48 GB usable — headroom for context

48GB · Ampere · model file 0.3GB · 82M

~1200–1900 tok/s

computed roofline: 768 GB/s × 0.46–0.76 efficiency window / 0.3GB (Q4_K_M) — bands, never points.

Run kokoro 82M on instead

Raspberry Pi 5 + AI Kit (Hailo-8L) — cheapest catalog machine that runs kokoro 82M fully ($110)

RTX 4090 (24GB, used) — upgrade path machine

every model the NVIDIA RTX A6000 can run — full list

explore the full catalog84 machines indexed · live prices · what each one can run
2026-09-17 · ← all models on the NVIDIA RTX A6000 · fit = weights + 4k KV vs usable memory