LightOnOCR-3-0.8B, developed by lightonai, is the lightest and most efficient model in the LightOnOCR-3 family, designed for end-to-end Optical Character Recognition (OCR) and comprehensive document understanding. Built on the Qwen3.5 vision-language architecture, it streamlines complex document pipelines into a single model featuring two primary modes: a default transcription mode that outputs full-page Markdown text, and a "grounding" mode that returns bounding box coordinates and semantic labels for every visual element. Beyond standard text extraction, the model provides advanced visual understanding by automatically describing images and converting charts into structured HTML data tables. Capable of processing diverse formats—including tables, receipts, forms, multi-column layouts, and math notation—LightOnOCR-3-0.8B offers a fast, versatile, and Apache 2.0-licensed solution for both commercial and research document processing tasks.
[!NOTE] MTP — Speculative Decoding: Yes | GGUF — LightOnOCR-3-0.8B-MTP(GGUF)
[!NOTE] Multi-Token Prediction (MTP) GGUF is a specialized GGUF model file format extension that integrates speculative decoding directly into the model weights to significantly accelerate local inference. Unlike traditional speculative decoding which requires a separate, smaller "draft" model, MTP GGUF files include additional output heads within the main model architecture that predict multiple future tokens in a single forward pass.
| File Name | Quant Type | File Size | File Link | Description |
|---|---|---|---|---|
| LightOnOCR-3-0.8B-MTP.BF16.gguf | BF16 | 1.56 GB | Link | Full BF16 weights. Highest quality, largest file size. |
| LightOnOCR-3-0.8B-MTP.Q3_K_L.gguf | Q3_K_L | 502 MB | Link | Lower quality but usable, good for low RAM availability. |
| LightOnOCR-3-0.8B-MTP.Q3_K_M.gguf | Q3_K_M | 476 MB | Link | Low quality. |
| LightOnOCR-3-0.8B-MTP.Q4_K_M.gguf | Q4_K_M | 542 MB | Link | Good quality, default size for most use cases, recommended. |
| LightOnOCR-3-0.8B-MTP.Q4_K_S.gguf | Q4_K_S | 517 MB | Link | Slightly lower quality with more space savings, recommended. |
| LightOnOCR-3-0.8B-MTP.Q5_K_M.gguf | Q5_K_M | 593 MB | Link | High quality, recommended. |
| LightOnOCR-3-0.8B-MTP.Q5_K_S.gguf | Q5_K_S | 578 MB | Link | High quality, recommended. |
| LightOnOCR-3-0.8B-MTP.Q6_K.gguf | Q6_K | 647 MB | Link | Very high quality, near perfect, recommended. |
| LightOnOCR-3-0.8B-MTP.Q8_0.gguf | Q8_0 | 834 MB | Link | Extremely high quality, generally unneeded but max available quant. |
| LightOnOCR-3-0.8B-MTP.mmproj-bf16.gguf | mmproj-bf16 | 207 MB | Link | Multimodal projection file in BF16 format. Used for vision/OCR capabilities. |
| File Name | File Type | File Size | File Link | Description |
|---|---|---|---|---|
| chat_template.jinja | Jinja Template | 7.76 kB | Link | Chat template for formatting conversations and prompts. |
| config.json | JSON | 2.69 kB | Link | Core model architecture and configuration parameters. |
| generation_config.json | JSON | 205 Bytes | Link | Default text generation parameters (e.g., temperature, top_p). |
| merges.txt | Text | 3.35 MB | Link | Byte-Pair Encoding (BPE) merge rules for the tokenizer. |
| model-00001-of-00003.safetensors | Safetensors | 696 MB | Link | Model weight shard 1/3. Required to load the model. |
| model-00002-of-00003.safetensors | Safetensors | 694 MB | Link | Model weight shard 2/3. Required to load the model. |
| model-00003-of-00003.safetensors | Safetensors | 358 MB | Link | Model weight shard 3/3. Required to load the model. |
| model.safetensors.index.json | JSON | 45 kB | Link | Index file mapping model weights to their respective shards. |
| preprocessor_config.json | JSON | 390 Bytes | Link | Configuration for preprocessing vision/image inputs. |
| processor_config.json | JSON | 1.19 kB | Link | Configuration for the multimodal processor (text + vision). |
| tokenizer.json | JSON | 20 MB | Link | Complete tokenizer vocabulary and encoding rules. |
| tokenizer_config.json | JSON | 1.17 kB | Link | Tokenizer-specific configuration parameters. |
| vocab.json | JSON | 6.72 MB | Link | Tokenizer vocabulary mapping tokens to IDs. |
LLM inference in C/C++ — https://github.com/ggml-org/llama.cpp
MTP-repackaged and quantized using llama.cpp by — prithivMLmods.
model-00001-of-00003.safetensors 696 MBmodel-00002-of-00003.safetensors 694 MBmodel-00003-of-00003.safetensors 358 MBHave the files? Submit a torrent.
Submit a torrentHave the complete files and permission to share them? Replace the example folder path. The script checks the files, makes a torrent and submits its magnet and checksums. Keep your torrent client seeding. Get your community key from your settings, and read the script before running it.
curl -fsSL https://pirateface.co/package.sh | bash -s -- --repo prithivMLmods/LightOnOCR-3-0.8B-MTP-Repacked ./model-folderlightonai/LightOnOCR-3-0.8B