Desktop GPU · NVIDIA

What AI models can a GeForce GTX 1650 run?

With 4 GB to work with, the question is not which model is best but which ones fit at all. The answer is: small ones, and they are better than they used to be. Bandwidth is the weak point at 128 GB/s. Models fit, then generate slowly, because every token means reading the whole active model out of memory.

Memory4 GB

Bandwidth128 GB/s

Usable for a model3.2 GB

Runtime backendCUDA

ArchitectureTuring

Power75 W

The short answer

Assuming an 8k context window and default settings, these are the models worth downloading first.

Best all-rounder

Qwen3.5 4B

IQ4_XS · 2.96 GB · about 36.2 tokens/s

Best for code

Starcoder2 3B

Q6_K · 2.69 GB · about 40.7 tokens/s

Largest that still runs well

Qwen3.5 4B

4.66B parameters · IQ4_XS · about 36.2 tokens/s

Fastest useful answer

Starcoder2 3B

about 40.7 tokens/s · 2.69 GB

See how fast it feels

GeForce GTX 1650 running Qwen3.5 4B at IQ4_XS

Wait for the first word2.1 s

Then writes at36.2 tok/s

Whole answer6.9 s

YouWhy does my model use more memory when the conversation gets longer?

Model

Simulated from our estimate at a 512-token question, not a recording. Assumes nothing else is competing for the GPU.

Where it sits

Our indicative price for this card is $100 used, reviewed 2026-09-01. That is 4.0 GB of memory and 128 GB/s of bandwidth for every $100. Prices move, so treat these as a place to start a search and not a quote.

Devices in a similar price band, fastest memory first:

Device	Price	Memory	Bandwidth	Best all-round pick
GeForce GTX 1660 SUPER	$110 used	6 GB	336 GB/s	Qwen3.5 9B, 54.1 t/s

Every model, scored on this device

Each row uses the highest-quality quantization that both fits and stays conversational. Speed is a single-stream estimate at 8k context.

Model	Params	Verdict	Download	Memory used	Speed	Max context
Gemma 4 E2B Instruct Gemma	5.12B	Just fits	IQ4_XS	2.91 GB	36.4 t/s	32k
Qwen3.5 4B Qwen	4.66B	Just fits	IQ4_XS	2.96 GB	36.2 t/s	8k
Gemma 3 4B Instruct Gemma	4.3B	Just fits	Q4_K_M	3.07 GB	34.7 t/s	8k
Qwen2.5 3B Instruct Qwen	3.09B	Just fits	Q6_K	2.98 GB	35.7 t/s	8k
LFM2.5 2.6B Liquid	2.7B	Just fits	Q8_0	3.14 GB	33.7 t/s	8k
Granite 4.1 3B Granite	3.4B	Just fits	Q4_K_M	2.89 GB	37.1 t/s	8k
Qwen3.5 2B Qwen	2.27B	Just fits	Q8_0	2.7 GB	39.9 t/s	32k
Llama 3.2 3B Instruct Llama	3.21B	Just fits	Q4_K_M	3.05 GB	35.1 t/s	8k
Phi 3 Mini 4k Instruct Phi	3.82B	Just fits	IQ4_XS	2.96 GB	36.3 t/s	4k
Phi 4 Mini Instruct Phi	3.84B	Just fits	Q3_K_M	3.12 GB	34.3 t/s	8k
Gemma 2 2B Instruct Gemma	2.61B	Just fits	Q6_K	2.95 GB	36.2 t/s	8k
Starcoder2 3B StarCoder	3.03B	Just fits	Q6_K	2.69 GB	40.7 t/s	16k
Qwen3 1.7B Qwen	2.03B	Just fits	Q6_K	2.76 GB	38.8 t/s	8k
Qwen3 4B Qwen	4.02B	Fits only at 2 to 3 bits	IQ3_XXS !	2.91 GB	36.8 t/s	8k
Qwen3.5 9B Qwen	9.65B	Fits only at 2 to 3 bits	IQ2_XXS !	2.78 GB	39.8 t/s	16k
DeepSeek R1 Distill Qwen 1.5B Qwen	1.78B	Runs great	Q8_0	2.07 GB	53.8 t/s	32k
Gemma 4 E4B Instruct Gemma	8B	Fits only at 2 to 3 bits	IQ2_XXS !	2.42 GB	45.6 t/s	32k
Qwen2.5 1.5B Instruct Qwen	1.54B	Runs great	Q8_0	2.06 GB	54 t/s	32k
Qwen2.5 VL 7B Instruct Qwen	8.29B	Fits only at 2 to 3 bits	IQ2_XXS !	2.68 GB	41 t/s	16k
SmolLM2 1.7B SmolLM	1.71B	Just fits	Q6_K	3.14 GB	33.6 t/s	8k
OLMo 2 0425 1B OLMo	1.48B	Just fits	Q8_0	2.6 GB	41.6 t/s	4k
Qwen2.5 7B Instruct Qwen	7.62B	Fits only at 2 to 3 bits	IQ2_XXS !	2.52 GB	44.1 t/s	16k
Qwen2.5 Coder 7B Instruct Qwen	7.62B	Fits only at 2 to 3 bits	IQ2_XXS !	2.52 GB	44.1 t/s	16k
Llama 3.2 1B Instruct Llama	1.24B	Runs great	Q8_0	1.82 GB	63.8 t/s	32k
Internlm3 8B Instruct InternLM	8.8B	Fits only at 2 to 3 bits	IQ2_XXS !	2.77 GB	39.9 t/s	16k
Mistral 7B Instruct V0.3 Mistral	7.25B	Fits only at 2 to 3 bits	IQ2_XXS !	3.11 GB	34.8 t/s	8k
LFM2.5 1.2B Instruct Liquid	1.17B	Runs great	Q8_0	1.46 GB	84.2 t/s	64k
Mistral 7B Instruct V0.2 Mistral	7.24B	Fits only at 2 to 3 bits	IQ2_XXS !	3.11 GB	34.8 t/s	8k
Gemma 3 1B Instruct Gemma	1B	Runs great	Q8_0	1.35 GB	90.7 t/s	32k
Qwen3.5 0.8B Qwen	0.87B	Runs great	Q8_0	1.28 GB	96.8 t/s	128k

Show the other 56 models

Qwen3 0.6B	0.75B	Runs great	Q8_0	1.92 GB	58.2 t/s	16k
Qwen2.5 0.5B Instruct	0.49B	Runs great	Q8_0	0.88 GB	162.8 t/s	32k
LFM2.5 350M	0.35B	Runs great	Q8_0	0.68 GB	252 t/s	64k
LFM2.5 8B A1B	8.47B	Partial offload	n/a	5.19 GB	46.2 t/s	n/a
Qwen3.5 35B A3B	35.95B	Partial offload	n/a	20.48 GB	27.2 t/s	n/a
DeepSeek Coder V2 Lite Instruct	15.71B	Partial offload	n/a	9.29 GB	25.6 t/s	n/a
Qwen3 Coder Next	79.67B	Partial offload	n/a	45.22 GB	21.4 t/s	n/a
Qwen3 Next 80B A3B Instruct	81.32B	Partial offload	n/a	46.15 GB	21.2 t/s	n/a
GPT OSS 20B	20.91B	Partial offload	n/a	11.55 GB	18.1 t/s	n/a
GLM 4.7 Flash	31.22B	Partial offload	n/a	18.13 GB	17.8 t/s	n/a
Qwen3 30B A3B	30.53B	Partial offload	n/a	18.08 GB	17 t/s	n/a
Qwen3 Coder 30B A3B Instruct	30.53B	Partial offload	n/a	18.08 GB	17 t/s	n/a
Gemma 4 26B A4B Instruct	25.81B	Partial offload	n/a	15.7 GB	15.3 t/s	n/a
GPT OSS 120B	116.83B	Partial offload	n/a	59.4 GB	12.6 t/s	n/a
Qwen3.5 122B A10B	125.09B	Partial offload	n/a	57.3 GB	10.7 t/s	n/a
Llama 3.1 8B Instruct	8.03B	Partial offload	n/a	5.63 GB	9 t/s	n/a
Qwen3 8B	8.19B	Partial offload	n/a	5.79 GB	8.4 t/s	n/a
Phi 3.5 MoE Instruct	41.87B	Partial offload	n/a	24.87 GB	7.9 t/s	n/a
Granite 4.1 8B	8.79B	Partial offload	n/a	6.6 GB	7 t/s	n/a
OLMo 3 7B Instruct	7.3B	Partial offload	n/a	6.78 GB	6.7 t/s	n/a
Gemma 2 9B Instruct	9.24B	Partial offload	n/a	7.55 GB	5.6 t/s	n/a
DeepSeek Coder 7B Instruct V1.5	6.91B	Partial offload	n/a	7.81 GB	5.4 t/s	n/a
Mistral Nemo Instruct 2407	12.25B	Partial offload	n/a	8.2 GB	5.1 t/s	n/a
Gemma 3 12B Instruct	12.19B	Partial offload	n/a	8.07 GB	5.1 t/s	n/a
DeepSeek Coder 6.7B Instruct	6.74B	Partial offload	n/a	8.12 GB	5.1 t/s	n/a
Gemma 4 12B Instruct	11.96B	Partial offload	n/a	8.44 GB	4.8 t/s	n/a
Mixtral 8x7B Instruct V0.1	46.7B	Partial offload	n/a	27.59 GB	4.1 t/s	n/a
Qwen3 14B	14.77B	Partial offload	n/a	9.56 GB	4.1 t/s	n/a
Qwen2.5 14B Instruct	14.77B	Partial offload	n/a	9.81 GB	4 t/s	n/a
Qwen2.5 Coder 14B Instruct	14.77B	Partial offload	n/a	9.81 GB	4 t/s	n/a
Phi 4	14.66B	Partial offload	n/a	9.96 GB	3.9 t/s	n/a
MiniMax M2.5	228.7B	Partial offload	n/a	130.56 GB	3.7 t/s	n/a
MiniMax M2.7	228.69B	Partial offload	n/a	130.55 GB	3.7 t/s	n/a
GLM 4.5 Air	110.47B	Partial offload	n/a	63.61 GB	3.4 t/s	n/a
Mistral Small 24B Instruct 2501	23.57B	Partial offload	n/a	14.57 GB	2.4 t/s	n/a
Codestral 22B V0.1	22.25B	Partial offload	n/a	14.62 GB	2.4 t/s	n/a
Qwen3 235B A22B	235.09B	Partial offload	n/a	133.71 GB	2.2 t/s	n/a
Qwen3.5 27B	27.78B	Partial offload	n/a	15.99 GB	2.2 t/s	n/a
Gemma 3 27B Instruct	27.43B	Partial offload	n/a	16.96 GB	2 t/s	n/a
Granite 4.1 30B	28.87B	Partial offload	n/a	18.64 GB	1.8 t/s	n/a
Qwen3 32B	32.76B	Partial offload	n/a	20.42 GB	1.6 t/s	n/a
Qwen2.5 32B Instruct	32.76B	Partial offload	n/a	20.42 GB	1.6 t/s	n/a
Gemma 4 31B Instruct	31.27B	Partial offload	n/a	21.31 GB	1.6 t/s	n/a
Qwen2.5 Coder 32B Instruct	32.76B	Partial offload	n/a	20.42 GB	1.6 t/s	n/a
GLM 4.5	358.34B	Partial offload	n/a	204.37 GB	1.4 t/s	n/a
Llama 3.3 70B Instruct	70.55B	Partial offload	n/a	42.12 GB	0.8 t/s	n/a
Qwen2.5 VL 72B Instruct	73.41B	Partial offload	n/a	43.62 GB	0.7 t/s	n/a
Qwen2.5 72B Instruct	72.71B	Partial offload	n/a	43.22 GB	0.7 t/s	n/a
Llama 3 3 Nemotron Super 49B V1	49.87B	Partial offload	n/a	47.99 GB	0.7 t/s	n/a
MiniMax M3	427.04B	Will not run	n/a	240.84 GB	1.9 t/s	n/a
GLM 5.1	753.86B	Will not run	n/a	424.51 GB	1.4 t/s	n/a
Kimi K2.6	1026.88B	Will not run	n/a	577.78 GB	1.3 t/s	n/a
Kimi K2 Instruct	1026.41B	Will not run	n/a	577.52 GB	1.3 t/s	n/a
DeepSeek V3.2	685.4B	Will not run	n/a	385.91 GB	1.3 t/s	n/a
DeepSeek R1	684.53B	Will not run	n/a	385.42 GB	1.3 t/s	n/a
GLM 5.2	753.33B	Will not run	n/a	424.21 GB	1 t/s	n/a

What it will not run

7 of the 86 architectures we track are out of reach here, even at 2-bit precision. Another 46 run by splitting the model between the card and system RAM, which works but drops generation to single digits for most dense models.

If you want more headroom

The next steps up in memory, in order. More memory changes which models load at all; more bandwidth changes how fast they answer, so the two are worth weighing separately.

Device	Memory	Bandwidth	Models that fit
GeForce GTX 1660 SUPER	6 GB	336 GB/s	43	Check price
GeForce RTX 3050 6GB	6 GB	168 GB/s	43	Check price
GeForce RTX 3070 Ti	8 GB	608 GB/s	49	Check price
Arc A750	8 GB	512 GB/s	49	Check price

Price links go to an Amazon search for the model name and are affiliate links: if you buy through one, we earn a commission at no cost to you. We do not take payment for placement, and the ordering above is by memory capacity alone.

Questions people ask about a GTX 1650

How much VRAM does a GeForce GTX 1650 have?

The GeForce GTX 1650 has 4 GB of VRAM, of which roughly 3.2 GB is available to a model once the operating system takes its share.

Can a GeForce GTX 1650 run a 70B model?

Only partially. About 48 of 48 layers fit on the card and the rest run on the CPU, which brings generation down to roughly 21.4 tokens per second.

What is the best local model for a GeForce GTX 1650?

Qwen3.5 4B at IQ4_XS is the strongest all-round fit: 2.96 GB used, around 36.2 tokens per second at 8k context.