Q: What quantization of Yi 1.5 34B should I use on a NVIDIA GeForce GTX 1660 Ti?

For 6 GB VRAM on the NVIDIA GeForce GTX 1660 Ti, the Q8_0 variant is the best fit. Estimated ~3 tokens/sec on the Q8_0 quantization.

Q: How fast does Yi 1.5 34B run on NVIDIA GeForce GTX 1660 Ti?

Roughly 3 tokens/sec for Q8_0. Real speed depends on context length, backend (Ollama, llama.cpp, LM Studio), and KV cache size.

Q: What if NVIDIA GeForce GTX 1660 Ti is not enough for Yi 1.5 34B?

Consider upgrading to Apple M4 Pro (48 GB VRAM) which fits the recommended 44 GB target. Or pick a smaller quantization to stay on your current card.

Question 1

Can I run Yi 1.5 34B on a NVIDIA GeForce GTX 1660 Ti?

Accepted Answer

Sort of — NVIDIA GeForce GTX 1660 Ti can run Yi 1.5 34B (Q8_0) only by spilling layers to RAM. Generation will be slow. CPU + GPU hybrid — not enough VRAM (6 GB < 38 GB min), but 64 GB RAM is sufficient. Expect significantly slower inference.

Question 2

What quantization of Yi 1.5 34B should I use on a NVIDIA GeForce GTX 1660 Ti?

Accepted Answer

For 6 GB VRAM on the NVIDIA GeForce GTX 1660 Ti, the Q8_0 variant is the best fit. Estimated ~3 tokens/sec on the Q8_0 quantization.

Question 3

How fast does Yi 1.5 34B run on NVIDIA GeForce GTX 1660 Ti?

Accepted Answer

Roughly 3 tokens/sec for Q8_0. Real speed depends on context length, backend (Ollama, llama.cpp, LM Studio), and KV cache size.

Question 4

What if NVIDIA GeForce GTX 1660 Ti is not enough for Yi 1.5 34B?

Accepted Answer

Consider upgrading to Apple M4 Pro (48 GB VRAM) which fits the recommended 44 GB target. Or pick a smaller quantization to stay on your current card.

Quantization	File Size	Min VRAM	Rec VRAM	Context	Verdict	Estimated tok/s
Q4_K_M	20 GB	22 GB	24 GB	4K / 4K	Hybrid CPU+GPU	~4
Q8_0Best fit	36 GB	38 GB	44 GB	4K / 4K	Hybrid CPU+GPU	~3

Can I Run Yi 1.5 34B on NVIDIA GeForce GTX 1660 Ti?

Share this matchup

Every Yi 1.5 34B quantization on NVIDIA GeForce GTX 1660 Ti

Upgrade options that fit Yi 1.5 34B better