RunsOnMy ModelsGPUs & MacsAPI pricesVRAM calculatorNew

Can I run Qwen2.5 Coder 7B Instruct locally?

Qwen2.5 Coder 7B Instruct by Qwen has 7.6B parameters. At 4-bit with an 8K context it needs about 5.5 GB of GPU memory. It fits on 65 of 65 GPUs and Macs we track.

5.5 GBQ4_K_M · 8K context
RTX 4060smallest consumer GPU that fits
M1 16GBsmallest Mac that fits
32Kmax context
GQAattention type

GGUF Q8_0 file was 200% of the size expected from the parameter count, so this page uses size estimates instead of real file sizes.

Memory needed by quant and context length

GB of GPU memory = weights + KV cache (FP16) + runtime overhead. Weights use parameter count × bits per weight. How we calculate.

QuantQuality4K ctx8K ctx32K ctx
2-bit (Q2_K)rough3.6 GB3.8 GB5.1 GB
3-bit (Q3_K_M)usable4.4 GB4.6 GB5.9 GB
4-bit (Q4_K_M)good, most popular5.2 GB5.5 GB6.8 GB
5-bit (Q5_K_M)very good6.0 GB6.2 GB7.5 GB
6-bit (Q6_K)near lossless6.7 GB6.9 GB8.3 GB
8-bit (Q8_0)effectively lossless8.4 GB8.7 GB10.0 GB
16-bit (F16/BF16)original weights15.1 GB15.3 GB16.6 GB

Which GPUs and Macs run Qwen2.5 Coder 7B Instruct?

Best quality quant that fits at 8K context. "Tight" = within 10% of usable memory. Speed is a ceiling from memory bandwidth (real speed is typically 50–80% of it).

HardwareMemoryFits?Best quantNeedsMax tok/s (est.)
RTX 3060 12GB12 GBFitsQ8_08.7 GB44details
RTX 4060 Ti 16GB16 GBFitsQ8_08.7 GB36details
RTX 309024 GBFitsF1615.3 GB61details
RTX 409024 GBFitsF1615.3 GB66details
RTX 509032 GBFitsF1615.3 GB118details
RTX 5070 Ti16 GBFitsQ8_08.7 GB111details
RX 7900 XTX24 GBFitsF1615.3 GB63details
Mac M4 Pro 48GB48 GBFitsF1615.3 GB18details
Mac M4 Max 128GB128 GBFitsF1615.3 GB36details
Mac M3 Ultra 512GB512 GBFitsF1615.3 GB54details
DGX Spark128 GBFitsF1615.3 GB18details
RTX 407012 GBFitsQ8_08.7 GB62details
RTX 5060 Ti 16GB16 GBFitsQ8_08.7 GB55details
RTX 40608 GBFitsQ5_K_M6.2 GB50details
Arc B58012 GBFitsQ8_08.7 GB56details
RTX 3060 Ti8 GBFitsQ5_K_M6.2 GB83details
RTX 30708 GBFitsQ5_K_M6.2 GB83details
RTX 3080 10GB10 GBFitsQ6_K6.9 GB122details
RTX 3080 12GB12 GBFitsQ8_08.7 GB113details
RTX 3090 Ti24 GBFitsF1615.3 GB66details
RTX 4060 Ti 8GB8 GBFitsQ5_K_M6.2 GB53details
RTX 4070 Super12 GBFitsQ8_08.7 GB62details
RTX 4070 Ti12 GBFitsQ8_08.7 GB62details
RTX 4070 Ti Super16 GBFitsQ8_08.7 GB83details
RTX 408016 GBFitsQ8_08.7 GB89details
RTX 4080 Super16 GBFitsQ8_08.7 GB91details
RTX 50608 GBFitsQ5_K_M6.2 GB83details
RTX 507012 GBFitsQ8_08.7 GB83details
RTX 508016 GBFitsQ8_08.7 GB119details
2x RTX 30902×24 GBFitsF1615.3 GB61details
2x RTX 40902×24 GBFitsF1615.3 GB66details
2x RTX 50902×32 GBFitsF1615.3 GB118details
RTX A600048 GBFitsF1615.3 GB50details
RTX 6000 Ada48 GBFitsF1615.3 GB63details
RTX PRO 6000 Blackwell96 GBFitsF1615.3 GB118details
L40S48 GBFitsF1615.3 GB57details
A100 40GB40 GBFitsF1615.3 GB102details
A100 80GB80 GBFitsF1615.3 GB134details
H100 PCIe80 GBFitsF1615.3 GB131details
H100 SXM80 GBFitsF1615.3 GB220details
H200141 GBFitsF1615.3 GB315details
RX 76008 GBFitsQ5_K_M6.2 GB53details
RX 7700 XT12 GBFitsQ8_08.7 GB53details
RX 7800 XT16 GBFitsQ8_08.7 GB77details
RX 7900 GRE16 GBFitsQ8_08.7 GB71details
RX 7900 XT20 GBFitsF1615.3 GB53details
RX 907016 GBFitsQ8_08.7 GB79details
RX 9070 XT16 GBFitsQ8_08.7 GB79details
Ryzen AI Max+ 395 128GB128 GBFitsF1615.3 GB17details
Arc A770 16GB16 GBFitsQ8_08.7 GB69details
Mac M1 16GB16 GBFitsQ8_08.7 GB8.4details
Mac M2 24GB24 GBFitsQ8_08.7 GB12details
Mac M3 24GB24 GBFitsQ8_08.7 GB12details
Mac M4 16GB16 GBFitsQ8_08.7 GB15details
Mac M4 32GB32 GBFitsF1615.3 GB7.9details
Mac M1 Pro 32GB32 GBFitsF1615.3 GB13details
Mac M3 Pro 36GB36 GBFitsF1615.3 GB9.8details
Mac M4 Pro 24GB24 GBFitsQ8_08.7 GB34details
Mac M4 Pro 64GB64 GBFitsF1615.3 GB18details
Mac M1 Max 64GB64 GBFitsF1615.3 GB26details
Mac M2 Max 96GB96 GBFitsF1615.3 GB26details
Mac M3 Max 128GB128 GBFitsF1615.3 GB26details
Mac M4 Max 64GB64 GBFitsF1615.3 GB36details
Mac M2 Ultra 192GB192 GBFitsF1615.3 GB53details
Mac M3 Ultra 256GB256 GBFitsF1615.3 GB54details

Questions

How much VRAM does Qwen2.5 Coder 7B Instruct need?

About 5.5 GB at 4-bit (Q4_K_M) with an 8K context: 4.3 GB of weights (estimated), 0.4 GB of KV cache and 0.7 GB of runtime overhead. At 8-bit it needs about 8.7 GB.

What is the smallest GPU that runs Qwen2.5 Coder 7B Instruct?

At Q4_K_M and 8K context, the smallest single consumer GPU in our table that fits it is the RTX 4060 (8 GB).

Can I run Qwen2.5 Coder 7B Instruct on a Mac?

Yes. At Q4_K_M with 8K context it fits on a Mac M1 16GB or larger, using the default macOS GPU memory limit.

Model facts

Hugging Face repoQwen/Qwen2.5-Coder-7B-Instruct
Parameters7,615,616,512 (7.6B)
Layers / hidden size28 / 3584
AttentionGQA; 28 heads, 4 KV heads, head dim 128
KV cache per token (FP16)56.0 KiB
Max context32,768 tokens
Licenseapache-2.0
Released on HFSep 17, 2024
Downloads (30d) / likes1,527,608 / 873

Sources: Hugging Face model API · config.json (Qwen/Qwen2.5-Coder-7B-Instruct) · GGUF files · GPU specs from vendor pages (list). Data fetched Nov 17, 2026. Spot an error? Report it.