local intelligence

can it run

Can the NVIDIA RTX 2060 run WHISPER small?

FULL

2.9 GB needed of 6 GB usable — headroom for context

6GB · Turing TU106 · model file 0.9GB · small

~180–270 tok/s

computed roofline: 336 GB/s × 0.48–0.72 efficiency window / 0.9GB (Q4_K_M) — bands, never points.

Run WHISPER small on instead

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

rtx-5060-ti-16 — the next step up from NVIDIA RTX 2060

every model the NVIDIA RTX 2060 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 2060 · fit = weights + 4k KV vs usable memory