RunsOnMy ModelsGPUs & MacsAPI pricesVRAM calculatorNew

Can I run GOT OCR2 0 locally?

GOT OCR2 0 by stepfun-ai has 716M parameters. At 4-bit with an 8K context it needs about 1.8 GB of GPU memory. It fits on 65 of 65 GPUs and Macs we track.

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

GGUF Q8_0 file was 158% 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)rough1.3 GB1.7 GB3.9 GB
3-bit (Q3_K_M)usable1.4 GB1.8 GB4.0 GB
4-bit (Q4_K_M)good, most popular1.5 GB1.8 GB4.1 GB
5-bit (Q5_K_M)very good1.5 GB1.9 GB4.2 GB
6-bit (Q6_K)near lossless1.6 GB2.0 GB4.2 GB
8-bit (Q8_0)effectively lossless1.8 GB2.1 GB4.4 GB
16-bit (F16/BF16)original weights2.4 GB2.8 GB5.0 GB

Which GPUs and Macs run GOT OCR2 0?

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 GBFitsF162.8 GB251details
RTX 4060 Ti 16GB16 GBFitsF162.8 GB201details
RTX 309024 GBFitsF162.8 GB654details
RTX 409024 GBFitsF162.8 GB704details
RTX 509032 GBFitsF162.8 GB1251details
RTX 5070 Ti16 GBFitsF162.8 GB626details
RX 7900 XTX24 GBFitsF162.8 GB670details
Mac M4 Pro 48GB48 GBFitsF162.8 GB191details
Mac M4 Max 128GB128 GBFitsF162.8 GB381details
Mac M3 Ultra 512GB512 GBFitsF162.8 GB572details
DGX Spark128 GBFitsF162.8 GB191details
RTX 407012 GBFitsF162.8 GB352details
RTX 5060 Ti 16GB16 GBFitsF162.8 GB313details
RTX 40608 GBFitsF162.8 GB190details
Arc B58012 GBFitsF162.8 GB318details
RTX 3060 Ti8 GBFitsF162.8 GB313details
RTX 30708 GBFitsF162.8 GB313details
RTX 3080 10GB10 GBFitsF162.8 GB531details
RTX 3080 12GB12 GBFitsF162.8 GB637details
RTX 3090 Ti24 GBFitsF162.8 GB704details
RTX 4060 Ti 8GB8 GBFitsF162.8 GB201details
RTX 4070 Super12 GBFitsF162.8 GB352details
RTX 4070 Ti12 GBFitsF162.8 GB352details
RTX 4070 Ti Super16 GBFitsF162.8 GB469details
RTX 408016 GBFitsF162.8 GB501details
RTX 4080 Super16 GBFitsF162.8 GB514details
RTX 50608 GBFitsF162.8 GB313details
RTX 507012 GBFitsF162.8 GB469details
RTX 508016 GBFitsF162.8 GB670details
2x RTX 30902×24 GBFitsF162.8 GB654details
2x RTX 40902×24 GBFitsF162.8 GB704details
2x RTX 50902×32 GBFitsF162.8 GB1251details
RTX A600048 GBFitsF162.8 GB536details
RTX 6000 Ada48 GBFitsF162.8 GB670details
RTX PRO 6000 Blackwell96 GBFitsF162.8 GB1251details
L40S48 GBFitsF162.8 GB603details
A100 40GB40 GBFitsF162.8 GB1086details
A100 80GB80 GBFitsF162.8 GB1424details
H100 PCIe80 GBFitsF162.8 GB1397details
H100 SXM80 GBFitsF162.8 GB2339details
H200141 GBFitsF162.8 GB3352details
RX 76008 GBFitsF162.8 GB201details
RX 7700 XT12 GBFitsF162.8 GB302details
RX 7800 XT16 GBFitsF162.8 GB436details
RX 7900 GRE16 GBFitsF162.8 GB402details
RX 7900 XT20 GBFitsF162.8 GB559details
RX 907016 GBFitsF162.8 GB447details
RX 9070 XT16 GBFitsF162.8 GB447details
Ryzen AI Max+ 395 128GB128 GBFitsF162.8 GB179details
Arc A770 16GB16 GBFitsF162.8 GB391details
Mac M1 16GB16 GBFitsF162.8 GB47details
Mac M2 24GB24 GBFitsF162.8 GB70details
Mac M3 24GB24 GBFitsF162.8 GB70details
Mac M4 16GB16 GBFitsF162.8 GB84details
Mac M4 32GB32 GBFitsF162.8 GB84details
Mac M1 Pro 32GB32 GBFitsF162.8 GB140details
Mac M3 Pro 36GB36 GBFitsF162.8 GB105details
Mac M4 Pro 24GB24 GBFitsF162.8 GB191details
Mac M4 Pro 64GB64 GBFitsF162.8 GB191details
Mac M1 Max 64GB64 GBFitsF162.8 GB279details
Mac M2 Max 96GB96 GBFitsF162.8 GB279details
Mac M3 Max 128GB128 GBFitsF162.8 GB279details
Mac M4 Max 64GB64 GBFitsF162.8 GB381details
Mac M2 Ultra 192GB192 GBFitsF162.8 GB559details
Mac M3 Ultra 256GB256 GBFitsF162.8 GB572details

Questions

How much VRAM does GOT OCR2 0 need?

About 1.8 GB at 4-bit (Q4_K_M) with an 8K context: 0.4 GB of weights (estimated), 0.8 GB of KV cache and 0.7 GB of runtime overhead. At 8-bit it needs about 2.1 GB.

What is the smallest GPU that runs GOT OCR2 0?

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 GOT OCR2 0 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 repostepfun-ai/GOT-OCR2_0
Parameters716,033,280 (716M)
Layers / hidden size24 / 1024
AttentionMHA; 16 heads, 16 KV heads, head dim 64
KV cache per token (FP16)96.0 KiB
Max context32,768 tokens
Licenseapache-2.0
Released on HFSep 12, 2024
Downloads (30d) / likes129,731 / 1,577

Sources: Hugging Face model API · config.json (stepfun-ai/GOT-OCR2_0) · GGUF files · GPU specs from vendor pages (list). Data fetched Nov 17, 2026. Spot an error? Report it.