Qwen on NVIDIA Blackwell

Can I run Qwen3 4B on an RTX 5060 Ti 16GB?

Yes. Qwen3 4B at Q4_K_M uses 4.2 GB of the 14.2 GB available on a RTX 5060 Ti 16GB, and runs at about 101 tokens per second. You can push the context to 79K.

Runs comfortably

4.2 GB of 14.2 GB · 30%
014 GB
Weights 2.3 GB
KV cache 1.1 GB
Runtime overhead 0.8 GB

Qwen3 4B at Q4_K_M leaves 10.0 GB spare on a RTX 5060 Ti 16GB. There is room to raise the context length or move up a quantisation level.

Generation101tok/s
Prompt processing2455tok/s
Max context79Ktokens
KV per 1K tokens0GB

Every quantisation of Qwen3 4B on a RTX 5060 Ti 16GB

Highlighted row is the highest quality that still fits at 8K context.

QuantisationbpwWeights TotalVerdictMax ctxtok/s
FP16 / BF16 16.00 7.5 GB 9.4 GB Runs comfortably 42K 36.5
INT8 / W8A8 8.50 3.9 GB 5.8 GB Runs comfortably 68K 65.6
Q8_0 (GGUF) 8.50 3.9 GB 5.8 GB Runs comfortably 68K 65.6
FP8 (E4M3) 8.00 3.7 GB 5.7 GB Runs comfortably 69K 67.8
Q6_K 6.56 3.1 GB 5.0 GB Runs comfortably 74K 80.1
Q5_K_M 5.67 2.7 GB 4.6 GB Runs comfortably 77K 90.3
Q5_K_S 5.52 2.6 GB 4.5 GB Runs comfortably 77K 92.3
AWQ 4-bit 4.25 2.5 GB 4.4 GB Runs comfortably 78K 94.1
GPTQ 4-bit 4.25 2.5 GB 4.4 GB Runs comfortably 78K 94.1
MXFP4 4.25 2.5 GB 4.4 GB Runs comfortably 78K 94.1
Q4_K_M 4.85 2.3 GB 4.2 GB Runs comfortably 79K 101
Q4_K_S 4.58 2.2 GB 4.1 GB Runs comfortably 80K 105
Q4_0 4.55 2.2 GB 4.1 GB Runs comfortably 80K 106
IQ4_XS 4.25 2.0 GB 4.0 GB Runs comfortably 81K 111
Q3_K_M 3.91 1.9 GB 3.8 GB Runs comfortably 82K 117
IQ3_M 3.70 1.8 GB 3.7 GB Runs comfortably 83K 121
IQ3_XXS 3.06 1.5 GB 3.5 GB Runs comfortably 85K 137
Q2_K 2.63 1.4 GB 3.3 GB Runs comfortably 86K 149
IQ2_XXS 2.06 1.1 GB 3.0 GB Runs comfortably 88K 170
IQ1_M 1.75 1.0 GB 2.9 GB Runs comfortably 89K 183

Also worth checking

Qwen3 4B on every GPU
Full hardware tableEverything that runs on a RTX 5060 Ti 16GB
Full model tableWhy the KV cache is the part that bites
Guide