local intelligence

can it run

Can the NVIDIA GTX 1080 run kokoro 82M?

FULL

2.3 GB needed of 8 GB usable — headroom for context

8GB · Pascal · model file 0.3GB · 82M

~300–530 tok/s

computed roofline: 320 GB/s × 0.28–0.50 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-5060-ti-16 — the next step up from NVIDIA GTX 1080

every model the NVIDIA GTX 1080 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 GTX 1080 · fit = weights + 4k KV vs usable memory