Nara Lightonocr 3 0.8b needs about 1.4 GB of memory at Q4_K_M (a 0.5 GB file) with an 8K context window, so it fits on a 4 GB GPU. This repository has 12 quants, from 0.4 GB (Q2_K) to 1.5 GB (F16). Longer context needs more: at 32K tokens the KV cache alone adds about 1.6 GB.
Line‑by‑line OCR for English archival documents
Nara Lightonocr 3 0.8b is a Qwen3.5 vision‑language model fine‑tuned for transcription of printed and handwritten English documents. It outputs one line per physical line in reading order. The model is trained on 38,756 pages of archival material.
Good at
- OCR of mixed printed and handwritten text
- Line‑by‑line transcription
- English‑only documents
Limits
- No vision support beyond OCR
- No multilingual capability
- Requires 8,192 token context for full pages
Best for
- Transcribing historical archives
- Digitizing pension files
- Processing catalogue cards
Who it’s for. Developers building OCR pipelines for English archival collections. Researchers needing a lightweight, open‑source model for line‑level transcription tasks.
Quants & memory
| Quant | File | Est. memory | vs VRAM | Result | |
|---|---|---|---|---|---|
| Q2_K | 0.4 GB | 1.3 GB | Fits | Get file → | |
| Q3_K_S | 0.4 GB | 1.3 GB | Fits | Get file → | |
| Q3_K_M | 0.5 GB | 1.4 GB | Fits | Get file → | |
| Q3_K_L | 0.5 GB | 1.4 GB | Fits | Get file → | |
| Q4_K_S | 0.5 GB | 1.4 GB | Fits | Get file → | |
| IQ4_XS | 0.5 GB | 1.4 GB | Fits | Get file → | |
| Q4_K_M | 0.5 GB | 1.4 GB | Fits | Get file → | |
| Q5_K_S | 0.6 GB | 1.5 GB | Fits | Get file → | |
| Q5_K_M | 0.6 GB | 1.5 GB | Fits | Get file → | |
| Q6_K | 0.6 GB | 1.5 GB | Fits | Get file → | |
| Q8_0 | 0.8 GB | 1.7 GB | Fits | Get file → | |
| F16 | 1.5 GB | 2.4 GB | Fits | Get file → |
For 8 GB VRAM, use Q4_K_M (file size 0.53 GB, estimated memory 1.4 GB). For 4 GB VRAM, Q3_K_M (0.47 GB, 1.4 GB). For 2 GB VRAM, Q2_K (0.42 GB, 1.3 GB).
What GPU do you need for Nara Lightonocr 3 0.8b?
These figures add the file size, an fp16 KV cache at an 8K context window (worked out from this model’s own header) and about 0.5 GB for the runtime. They are estimates, so leave a little headroom, and expect to use less context or a smaller quant if you are right at the limit.
- 4 GB VRAM: the best quant that fits is F16 (1.5 GB file, about 2.4 GB in use).
Which Nara Lightonocr 3 0.8b quant should you choose?
Q4_K_M is the usual starting point for Nara Lightonocr 3 0.8b: it balances file size and quality well. Here it is a 0.5 GB file.
The Q8_0 file is 0.8 GB, about 1.5 times the size of Q4_K_M. It keeps the weights closer to the original, which only matters if you have the memory to spare.
If memory is tight, the smallest file is Q2_K at 0.4 GB, with the quality trade-offs that come with fewer bits. Use it only when memory is very tight. Expect a clear drop in quality, which is usually worse on small models and easier to live with on large ones.
New to quant names? Read GGUF quantization explained.
Specifications
Architecture from the GGUF header
Source
Frequently asked questions about Nara Lightonocr 3 0.8b
How much VRAM does Nara Lightonocr 3 0.8b need?
About 1.4 GB at Q4_K_M with an 8K context window. That is the 0.5 GB file, plus the KV cache and roughly 0.5 GB of runtime overhead. Smaller quants need less and longer context needs more.
Can I run Nara Lightonocr 3 0.8b on an 8 GB GPU?
Yes, with F16: the 1.5 GB file needs about 2.4 GB at 8K context. Larger quants in this repository do not fit in 8 GB.
What is the file size of Nara Lightonocr 3 0.8b GGUF?
It depends on the quant: Q2_K is 0.4 GB, Q3_K_S is 0.4 GB, Q3_K_M is 0.5 GB, Q3_K_L is 0.5 GB, Q4_K_S is 0.5 GB, IQ4_XS is 0.5 GB and 6 more.
What context length does Nara Lightonocr 3 0.8b support?
The GGUF header lists a trained context length of 262,144 tokens (256K). The KV cache grows with the context you use: about 0.4 GB at 8K and 1.6 GB at 32K.
Where can I download Nara Lightonocr 3 0.8b GGUF?
The files are on Hugging Face in the mradermacher/nara-lightonocr-3-0.8b-GGUF repository. The download buttons on this page open the file on Hugging Face. ggufmodels.com does not host any model files.
What license does Nara Lightonocr 3 0.8b use?
Hugging Face lists the license as apache-2.0. Licenses can differ between the quantized files and the original model, so check the repository and the base model before commercial use.
Can I run this model on 4 GB VRAM?
Yes, the Q3_K_M quant uses approximately 1.4 GB of VRAM at 8,192 context tokens.
Does the model support other languages?
No, the model is trained only on English archival documents.
Is the model licensed for commercial use?
The model is released under the apache-2.0 license, allowing commercial use.