RunsOnMy ModelsGPUs & MacsBy VRAMAPI pricesVRAM calculatorNew

Can I run Qwen3.5 2B locally?

Qwen3.5 2B by Qwen has 2.3B parameters. At 4-bit with an 8K context it needs about 2.2 GB of GPU memory. It fits on 65 of 65 GPUs and Macs we track. See all models that fit in 8 GB of VRAM.

2.2 GBQ4_K_M · 8K context
RTX 4060smallest consumer GPU that fits
M1 16GBsmallest Mac that fits
256Kmax context
GQAattention type

Memory needed by quant and context length

GB of GPU memory = weights + KV cache (FP16) + runtime overhead. Weights use real file sizes from unsloth/Qwen3.5-2B-GGUF where available (marked •), otherwise parameter count × bits per weight. How we calculate.

QuantQuality4K ctx8K ctx32K ctx128K ctx
2-bit (Q2_K)rough1.7 GB1.8 GB2.0 GB3.2 GB
3-bit (Q3_K_M) •usable2.0 GB2.0 GB2.3 GB3.4 GB
4-bit (Q4_K_M) •good, most popular2.1 GB2.2 GB2.4 GB3.6 GB
5-bit (Q5_K_M) •very good2.3 GB2.3 GB2.6 GB3.7 GB
6-bit (Q6_K) •near lossless2.4 GB2.4 GB2.7 GB3.8 GB
8-bit (Q8_0) •effectively lossless2.8 GB2.8 GB3.1 GB4.2 GB
16-bit (F16/BF16) •original weights4.4 GB4.5 GB4.8 GB5.9 GB

Which GPUs and Macs run Qwen3.5 2B?

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 GBFitsF164.5 GB95details
RTX 4060 Ti 16GB16 GBFitsF164.5 GB76details
RTX 309024 GBFitsF164.5 GB248details
RTX 409024 GBFitsF164.5 GB267details
RTX 509032 GBFitsF164.5 GB475details
RTX 5070 Ti16 GBFitsF164.5 GB237details
RX 7900 XTX24 GBFitsF164.5 GB254details
Mac M4 Pro 48GB48 GBFitsF164.5 GB72details
Mac M4 Max 128GB128 GBFitsF164.5 GB145details
Mac M3 Ultra 512GB512 GBFitsF164.5 GB217details
DGX Spark128 GBFitsF164.5 GB72details
RTX 407012 GBFitsF164.5 GB133details
RTX 5060 Ti 16GB16 GBFitsF164.5 GB119details
RTX 40608 GBFitsF164.5 GB72details
Arc B58012 GBFitsF164.5 GB121details
RTX 3060 Ti8 GBFitsF164.5 GB119details
RTX 30708 GBFitsF164.5 GB119details
RTX 3080 10GB10 GBFitsF164.5 GB201details
RTX 3080 12GB12 GBFitsF164.5 GB242details
RTX 3090 Ti24 GBFitsF164.5 GB267details
RTX 4060 Ti 8GB8 GBFitsF164.5 GB76details
RTX 4070 Super12 GBFitsF164.5 GB133details
RTX 4070 Ti12 GBFitsF164.5 GB133details
RTX 4070 Ti Super16 GBFitsF164.5 GB178details
RTX 408016 GBFitsF164.5 GB190details
RTX 4080 Super16 GBFitsF164.5 GB195details
RTX 50608 GBFitsF164.5 GB119details
RTX 507012 GBFitsF164.5 GB178details
RTX 508016 GBFitsF164.5 GB254details
2x RTX 30902×24 GBFitsF164.5 GB248details
2x RTX 40902×24 GBFitsF164.5 GB267details
2x RTX 50902×32 GBFitsF164.5 GB475details
RTX A600048 GBFitsF164.5 GB203details
RTX 6000 Ada48 GBFitsF164.5 GB254details
RTX PRO 6000 Blackwell96 GBFitsF164.5 GB475details
L40S48 GBFitsF164.5 GB229details
A100 40GB40 GBFitsF164.5 GB412details
A100 80GB80 GBFitsF164.5 GB540details
H100 PCIe80 GBFitsF164.5 GB530details
H100 SXM80 GBFitsF164.5 GB887details
H200141 GBFitsF164.5 GB1271details
RX 76008 GBFitsF164.5 GB76details
RX 7700 XT12 GBFitsF164.5 GB114details
RX 7800 XT16 GBFitsF164.5 GB165details
RX 7900 GRE16 GBFitsF164.5 GB153details
RX 7900 XT20 GBFitsF164.5 GB212details
RX 907016 GBFitsF164.5 GB170details
RX 9070 XT16 GBFitsF164.5 GB170details
Ryzen AI Max+ 395 128GB128 GBFitsF164.5 GB68details
Arc A770 16GB16 GBFitsF164.5 GB148details
Mac M1 16GB16 GBFitsF164.5 GB18details
Mac M2 24GB24 GBFitsF164.5 GB26details
Mac M3 24GB24 GBFitsF164.5 GB26details
Mac M4 16GB16 GBFitsF164.5 GB32details
Mac M4 32GB32 GBFitsF164.5 GB32details
Mac M1 Pro 32GB32 GBFitsF164.5 GB53details
Mac M3 Pro 36GB36 GBFitsF164.5 GB40details
Mac M4 Pro 24GB24 GBFitsF164.5 GB72details
Mac M4 Pro 64GB64 GBFitsF164.5 GB72details
Mac M1 Max 64GB64 GBFitsF164.5 GB106details
Mac M2 Max 96GB96 GBFitsF164.5 GB106details
Mac M3 Max 128GB128 GBFitsF164.5 GB106details
Mac M4 Max 64GB64 GBFitsF164.5 GB145details
Mac M2 Ultra 192GB192 GBFitsF164.5 GB212details
Mac M3 Ultra 256GB256 GBFitsF164.5 GB217details

Questions

How much VRAM does Qwen3.5 2B need?

About 2.2 GB at 4-bit (Q4_K_M) with an 8K context: 1.2 GB of weights (real GGUF file size), 0.1 GB of KV cache and 0.9 GB of runtime overhead. At 8-bit it needs about 2.8 GB.

What is the smallest GPU that runs Qwen3.5 2B?

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 Qwen3.5 2B 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.

Similar models

Same family or a similar parameter count, with the VRAM each needs at Q4_K_M and 8K context.

Qwen3 1.7B

2.0B · 2.6 GB at 4-bit

Qwen2.5 1.5B Instruct

1.5B · 1.9 GB at 4-bit

Qwen2.5 3B Instruct

3.1B · 2.9 GB at 4-bit

SmolVLM Instruct

2.2B · 3.0 GB at 4-bit

WeVisOne VL 2B

2.4B · 2.8 GB at 4-bit

MiniCPM5 2B

2.5B · 2.5 GB at 4-bit

Model facts

Hugging Face repoQwen/Qwen3.5-2B
Parameters2,274,069,824 (2.3B)
Layers / hidden size24 / 2048
AttentionGQA (hybrid: 18 linear/state layers); 8 heads, 2 KV heads, head dim 256
KV cache per token (FP16)12.0 KiB
Max context262,144 tokens
Licenseapache-2.0
Released on HFFeb 28, 2026
Downloads (30d) / likes4,915,024 / 487

Sources: Hugging Face model API · config.json (Qwen/Qwen3.5-2B) · GGUF files · GPU specs from vendor pages (list). Data fetched Nov 17, 2026. Spot an error? Report it.