can it run
4.6 GB vs 4 GB usable — partial CPU offload, expect large speed loss With a q8 (quantized) KV cache it comes to the edge (4.3GB) — possible with modest context, at some quality cost.
4GB · Turing · model file 4.1GB · 7B
computed roofline: 128.1 GB/s × 0.48–0.72 efficiency window / 4.1GB (Q4_K_M) — bands, never points.
→ Raspberry Pi 5 + AI Kit (Hailo-8L) — cheapest catalog machine that runs mistral 7B fully ($110)
→ gtx-1650-super — the next step up from NVIDIA GeForce GTX 1650
→ every model the NVIDIA GeForce GTX 1650 can run — full list
→explore the full catalog84 machines indexed · live prices · what each one can run