Nanonets OCR S — Hardware Requirements & GPU Compatibility
Vision
Nanonets-OCR-s is Nanonets' image-to-markdown OCR model, built on top of Qwen2.5-VL-3B-Instruct, that goes beyond plain text extraction to produce structured markdown for downstream processing by other language models. It converts mathematical equations and formulas into LaTeX, describes embedded images and charts inside structured tags, isolates signatures and watermarks into their own tags, converts checkboxes into standard Unicode symbols, and extracts complex tables into both markdown and HTML formats. At under 4 billion parameters it is light enough to run on a single consumer GPU. Context length is 128,000 tokens. Nanonets has not published a license for the model on its Hugging Face model card. It was published in June 2025.
209.2K downloads
1.6K likesSpecifications
- Publisher
- nanonets
- Parameters
- 3.8B
- Architecture
- Qwen2_5_VLForConditionalGeneration
- Context Length
- 128,000 tokens
- Vocabulary Size
- 151,936
- Release Date
- 2025-06-10
Get Started
HuggingFace
How Much VRAM Does Nanonets OCR S Need?
Select a quantization to see compatible GPUs below.
| Quantization | Bits | VRAM | + Context | File Size | Quality |
|---|---|---|---|---|---|
| Q2_K | 3.40 | 2.0 GB | 6.6 GB | 1.60 GB | 2-bit quantization with K-quant improvements |
| Q3_K_S | 3.50 | 2.0 GB | 6.7 GB | 1.64 GB | 3-bit small quantization |
| Q3_K_M | 3.90 | 2.2 GB | 6.8 GB | 1.83 GB | 3-bit medium quantization |
| Q4_0 | 4.00 | 2.3 GB | 6.9 GB | 1.88 GB | 4-bit legacy quantization |
| Q4_K_M | 4.80 | 2.6 GB | 7.3 GB | 2.25 GB | 4-bit medium quantization — most popular sweet spot |
| Q5_K_M | 5.70 | 3.0 GB | 7.7 GB | 2.68 GB | 5-bit medium quantization — good quality/size tradeoff |
| Q6_K | 6.60 | 3.5 GB | 8.1 GB | 3.10 GB | 6-bit quantization, very good quality |
| Q8_0 | 8.00 | 4.1 GB | 8.8 GB | 3.75 GB | 8-bit quantization, near-lossless |
Which GPUs Can Run Nanonets OCR S?
Q4_K_M · 2.6 GBNanonets OCR S (Q4_K_M) requires 2.6 GB of VRAM to load the model weights. For comfortable inference with headroom for KV cache and system overhead, 4+ GB is recommended. Using the full 128K context window can add up to 4.6 GB, bringing total usage to 7.3 GB. 52 GPUs can run it, including NVIDIA GeForce RTX 5090, NVIDIA GeForce RTX 3090 Ti.
Runs great
— Plenty of headroomWhich Devices Can Run Nanonets OCR S?
Q4_K_M · 2.6 GB59 devices with unified memory can run Nanonets OCR S, including NVIDIA DGX H100, NVIDIA DGX A100 640GB.
Runs great
— Plenty of headroomWhere to Download Nanonets OCR S
Community quantizations of this model — GGUF for llama.cpp, Ollama, and LM Studio, plus AWQ/MLX variants where available.
Related Models
Frequently Asked Questions
- How much VRAM does Nanonets OCR S need?
Nanonets OCR S requires 2.6 GB of VRAM at Q4_K_M, or 7.9 GB at BF16. Full 128K context adds up to 4.6 GB (7.3 GB total).
VRAM = Weights + KV Cache + Overhead
Weights = 3.8B × 4.8 bits ÷ 8 = 2.3 GB
KV Cache + Overhead ≈ 0.3 GB (at 2K context + ~0.3 GB framework)
KV Cache + Overhead ≈ 5 GB (at full 128K context)
VRAM usage by quantization
Q4_K_M2.6 GBQ4_K_M + full context7.3 GB- What's the best quantization for Nanonets OCR S?
For Nanonets OCR S, Q4_K_M (2.6 GB) offers the best balance of quality and VRAM usage. Q5_K_S (3.0 GB) provides better quality if you have the VRAM. The smallest option is IQ2_XXS at 1.4 GB.
VRAM requirement by quantization
IQ2_XXS1.4 GBQ3_K_S2.0 GBQ4_12.5 GBQ4_K_M ★2.6 GBQ5_K_S3.0 GBBF167.9 GB★ Recommended — best balance of quality and VRAM usage.
- Can I run Nanonets OCR S on a Mac?
Nanonets OCR S requires at least 1.4 GB at IQ2_XXS, which exceeds the unified memory of most consumer Macs. You would need a Mac Studio or Mac Pro with a high-memory configuration.
- Can I run Nanonets OCR S locally?
Yes — Nanonets OCR S can run locally on consumer hardware. At Q4_K_M quantization it needs 2.6 GB of VRAM. Popular tools include Ollama, LM Studio, and llama.cpp.
- How fast is Nanonets OCR S?
At Q4_K_M, Nanonets OCR S can reach ~1825 tok/s on AMD Instinct MI350X. On NVIDIA GeForce RTX 4090: ~249 tok/s. Speed depends mainly on GPU memory bandwidth. Real-world results typically within ±20%.
tok/s = (bandwidth GB/s ÷ model GB) × efficiency
Example: NVIDIA B200 → 8000 ÷ 2.6 × 0.65 = ~1977 tok/s
Estimated speed at Q4_K_M (2.6 GB)
~1977 tok/s~249 tok/s~1977 tok/s~1825 tok/sReal-world results typically within ±20%. Speed depends on batch size, quantization kernel, and software stack.
- What's the download size of Nanonets OCR S?
At Q4_K_M, the download is about 2.25 GB. The full-precision BF16 version is 7.51 GB. The smallest option (IQ2_XXS) is 1.03 GB.
- Which GPUs can run Nanonets OCR S?
52 consumer GPUs can run Nanonets OCR S at Q4_K_M (2.6 GB). Top options include AMD Radeon RX 6700 XT, AMD Radeon RX 6800, AMD Radeon RX 6800 XT. 52 GPUs have plenty of headroom for comfortable inference.
- Which devices can run Nanonets OCR S?
59 devices with unified memory can run Nanonets OCR S at Q4_K_M (2.6 GB), including AMD Ryzen AI 9 HX 370 (Strix Point) Laptop, ASUS Ascent GX10, Apple iPhone 17 Pro, Asus ROG Flow Z13 (2025, Ryzen AI Max+ 395, 128 GB). Apple Silicon Macs use unified memory shared between CPU and GPU, making them well-suited for local LLM inference.