Qwen on NVIDIA Ada

Can I run Qwen2.5 7B on an RTX 4070?

Yes. Qwen2.5 7B at Q4_K_M uses 5.6 GB of the 10.5 GB available on a RTX 4070, and runs at about 64.9 tokens per second. You can push the context to 97K.

Runs comfortably

5.6 GB of 10.5 GB · 54%
010 GB
Weights 4.4 GB
KV cache 0.4 GB
Runtime overhead 0.8 GB

Qwen2.5 7B at Q4_K_M leaves 4.8 GB spare on a RTX 4070. There is room to raise the context length or move up a quantisation level.

Generation64.9tok/s
Prompt processing1598tok/s
Max context97Ktokens
KV per 1K tokens0GB

Every quantisation of Qwen2.5 7B on a RTX 4070

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

QuantisationbpwWeights TotalVerdictMax ctxtok/s
FP16 / BF16 16.00 14.2 GB 15.4 GB Won't fit — 6.05
INT8 / W8A8 8.50 7.3 GB 8.5 GB Runs comfortably 43K 40.0
Q8_0 (GGUF) 8.50 7.3 GB 8.5 GB Runs comfortably 43K 40.0
FP8 (E4M3) 8.00 7.1 GB 8.3 GB Runs comfortably 47K 41.0
Q6_K 6.56 5.8 GB 7.1 GB Runs comfortably 70K 49.6
AWQ 4-bit 4.25 5.3 GB 6.5 GB Runs comfortably 81K 54.6
GPTQ 4-bit 4.25 5.3 GB 6.5 GB Runs comfortably 81K 54.6
MXFP4 4.25 5.3 GB 6.5 GB Runs comfortably 81K 54.6
Q5_K_M 5.67 5.0 GB 6.3 GB Runs comfortably 85K 57.0
Q5_K_S 5.52 4.9 GB 6.1 GB Runs comfortably 87K 58.5
Q4_K_M 4.85 4.4 GB 5.6 GB Runs comfortably 97K 64.9
Q4_K_S 4.58 4.2 GB 5.4 GB Runs comfortably 100K 67.9
Q4_0 4.55 4.2 GB 5.4 GB Runs comfortably 101K 68.3
IQ4_XS 4.25 3.9 GB 5.2 GB Runs comfortably 105K 72.0
Q3_K_M 3.91 3.7 GB 4.9 GB Runs comfortably 110K 76.7
IQ3_M 3.70 3.5 GB 4.8 GB Runs comfortably 113K 80.0
IQ3_XXS 3.06 3.0 GB 4.3 GB Runs comfortably 121K 91.8
Q2_K 2.63 2.7 GB 3.9 GB Runs comfortably 127K 102
IQ2_XXS 2.06 2.3 GB 3.5 GB Runs comfortably 128K 119
IQ1_M 1.75 2.0 GB 3.3 GB Runs comfortably 128K 132

Also worth checking

Qwen2.5 7B on every GPU
Full hardware tableEverything that runs on a RTX 4070
Full model tableWhy the KV cache is the part that bites
Guide