▲ local intelligence

can it run

Can the NVIDIA RTX Spark N1X run kokoro 82M?

FULL

2.3 GB needed of 22 GB usable — headroom for context

~22GB usable · NVIDIA RTX Spark N1X (Grace + Blackwell RTX) · model file 0.3GB · 82M

~230–590 tok/s

computed roofline: 273 GB/s × 0.25–0.65 efficiency window / 0.3GB (Q4_K_M) — bands, never points.

Run kokoro 82M on instead

→ Tesla M60 (16GB, used) — cheapest catalog machine that runs kokoro 82M fully ($35)

→ rtx-5090 — the next step up from NVIDIA RTX Spark N1X

→ every model the NVIDIA RTX Spark N1X can run — full list

→explore the full catalog97 machines indexed · live prices · what each one can run
2026-10-08 · ← all models on the NVIDIA RTX Spark N1X · fit = weights + 4k KV vs usable memory