RunsOnMy ModelsGPUs & MacsBy VRAMAPI pricesVRAM calculatorNew

Can I run PaddleOCR VL 1.6 locally?

PaddleOCR VL 1.6 by PaddlePaddle has 959M parameters. At 4-bit with an 8K context it needs about 1.3 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.

1.3 GBQ4_K_M · 8K context
RTX 4060smallest consumer GPU that fits
M1 16GBsmallest Mac that fits
128Kmax context
GQAattention type

GGUF Q8_0 file was 49% of the size expected from the parameter count, so this page uses size estimates instead of real file sizes.

This model uses a newer compressed or sparse attention design. Our KV-cache numbers use the standard formula, so treat long-context figures as an upper bound.

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 ctx128K ctx
2-bit (Q2_K)rough1.0 GB1.1 GB1.5 GB3.2 GB
3-bit (Q3_K_M)usable1.1 GB1.2 GB1.6 GB3.3 GB
4-bit (Q4_K_M)good, most popular1.2 GB1.3 GB1.7 GB3.4 GB
5-bit (Q5_K_M)very good1.3 GB1.4 GB1.8 GB3.5 GB
6-bit (Q6_K)near lossless1.4 GB1.5 GB1.9 GB3.6 GB
8-bit (Q8_0)effectively lossless1.6 GB1.7 GB2.1 GB3.8 GB
16-bit (F16/BF16)original weights2.5 GB2.5 GB2.9 GB4.6 GB

Which GPUs and Macs run PaddleOCR VL 1.6?

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 GBFitsF162.5 GB188details
RTX 4060 Ti 16GB16 GBFitsF162.5 GB150details
RTX 309024 GBFitsF162.5 GB488details
RTX 409024 GBFitsF162.5 GB526details
RTX 509032 GBFitsF162.5 GB935details
RTX 5070 Ti16 GBFitsF162.5 GB467details
RX 7900 XTX24 GBFitsF162.5 GB501details
Mac M4 Pro 48GB48 GBFitsF162.5 GB142details
Mac M4 Max 128GB128 GBFitsF162.5 GB285details
Mac M3 Ultra 512GB512 GBFitsF162.5 GB427details
DGX Spark128 GBFitsF162.5 GB142details
RTX 407012 GBFitsF162.5 GB263details
RTX 5060 Ti 16GB16 GBFitsF162.5 GB234details
RTX 40608 GBFitsF162.5 GB142details
Arc B58012 GBFitsF162.5 GB238details
RTX 3060 Ti8 GBFitsF162.5 GB234details
RTX 30708 GBFitsF162.5 GB234details
RTX 3080 10GB10 GBFitsF162.5 GB396details
RTX 3080 12GB12 GBFitsF162.5 GB476details
RTX 3090 Ti24 GBFitsF162.5 GB526details
RTX 4060 Ti 8GB8 GBFitsF162.5 GB150details
RTX 4070 Super12 GBFitsF162.5 GB263details
RTX 4070 Ti12 GBFitsF162.5 GB263details
RTX 4070 Ti Super16 GBFitsF162.5 GB351details
RTX 408016 GBFitsF162.5 GB374details
RTX 4080 Super16 GBFitsF162.5 GB384details
RTX 50608 GBFitsF162.5 GB234details
RTX 507012 GBFitsF162.5 GB351details
RTX 508016 GBFitsF162.5 GB501details
2x RTX 30902×24 GBFitsF162.5 GB488details
2x RTX 40902×24 GBFitsF162.5 GB526details
2x RTX 50902×32 GBFitsF162.5 GB935details
RTX A600048 GBFitsF162.5 GB401details
RTX 6000 Ada48 GBFitsF162.5 GB501details
RTX PRO 6000 Blackwell96 GBFitsF162.5 GB935details
L40S48 GBFitsF162.5 GB451details
A100 40GB40 GBFitsF162.5 GB811details
A100 80GB80 GBFitsF162.5 GB1064details
H100 PCIe80 GBFitsF162.5 GB1043details
H100 SXM80 GBFitsF162.5 GB1747details
H200141 GBFitsF162.5 GB2504details
RX 76008 GBFitsF162.5 GB150details
RX 7700 XT12 GBFitsF162.5 GB225details
RX 7800 XT16 GBFitsF162.5 GB325details
RX 7900 GRE16 GBFitsF162.5 GB300details
RX 7900 XT20 GBFitsF162.5 GB417details
RX 907016 GBFitsF162.5 GB334details
RX 9070 XT16 GBFitsF162.5 GB334details
Ryzen AI Max+ 395 128GB128 GBFitsF162.5 GB134details
Arc A770 16GB16 GBFitsF162.5 GB292details
Mac M1 16GB16 GBFitsF162.5 GB35details
Mac M2 24GB24 GBFitsF162.5 GB52details
Mac M3 24GB24 GBFitsF162.5 GB52details
Mac M4 16GB16 GBFitsF162.5 GB63details
Mac M4 32GB32 GBFitsF162.5 GB63details
Mac M1 Pro 32GB32 GBFitsF162.5 GB104details
Mac M3 Pro 36GB36 GBFitsF162.5 GB78details
Mac M4 Pro 24GB24 GBFitsF162.5 GB142details
Mac M4 Pro 64GB64 GBFitsF162.5 GB142details
Mac M1 Max 64GB64 GBFitsF162.5 GB209details
Mac M2 Max 96GB96 GBFitsF162.5 GB209details
Mac M3 Max 128GB128 GBFitsF162.5 GB209details
Mac M4 Max 64GB64 GBFitsF162.5 GB285details
Mac M2 Ultra 192GB192 GBFitsF162.5 GB417details
Mac M3 Ultra 256GB256 GBFitsF162.5 GB427details

Questions

How much VRAM does PaddleOCR VL 1.6 need?

About 1.3 GB at 4-bit (Q4_K_M) with an 8K context: 0.5 GB of weights (estimated), 0.1 GB of KV cache and 0.6 GB of runtime overhead. At 8-bit it needs about 1.7 GB.

What is the smallest GPU that runs PaddleOCR VL 1.6?

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 PaddleOCR VL 1.6 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.

Qwen3.5 0.8B

873M · 1.5 GB at 4-bit

LightOnOCR 3 0.8B

853M · 1.5 GB at 4-bit

HunyuanOCR

1.1B · 2.0 GB at 4-bit

FastVLM 0.5B

759M · 1.2 GB at 4-bit

Qwen3 0.6B

752M · 1.9 GB at 4-bit

LFM2.5 1.2B Instruct

1.2B · 1.3 GB at 4-bit

Model facts

Hugging Face repoPaddlePaddle/PaddleOCR-VL-1.6
Parameters958,588,736 (959M)
Layers / hidden size18 / 1024
AttentionGQA; 16 heads, 2 KV heads, head dim 128
KV cache per token (FP16)18.0 KiB
Max context131,072 tokens
Licenseapache-2.0
Released on HFMay 27, 2026
Downloads (30d) / likes98,553 / 593

Sources: Hugging Face model API · config.json (PaddlePaddle/PaddleOCR-VL-1.6) · GGUF files · GPU specs from vendor pages (list). Data fetched Nov 17, 2026. Spot an error? Report it.