This model was converted to MLX format from zai-org/GLM-OCR using mlx-vlm version 0.3.10. Refer to the original model card for more details on the model.
Full BF16 version of the model. We recommend this variant for inference and further fine-tuning. LightOnOCR-1B is a compact, end-to-end vision–language model for Optical Character Recognition (OCR) and document understanding.
Parameters1.2B
Context8,192
Weights2.3 GB
Licenseapache-2.0
AccessOpen weights
Monthly Downloads56k
Runs On
What it takes to serve LightOnOCR-1B-1025 (1.2B parameters): the memory its weights need at each precision, and the cheapest way to rent enough data-center GPUs to hold them.
| Precision | Weights | Memory needed | Cheapest setup | Per hour | Also fits |
|---|---|---|---|---|---|
| 16-bit | 2.3 GB | 2.8 GB | 1x MI300X (192 GB) Vultr |
$1.85 | 1x H100 $1.99 · 1x MI325X $2.00 |
| 8-bit | 1.2 GB | 1.4 GB | 1x MI300X (192 GB) Vultr |
$1.85 | 1x H100 $1.99 · 1x MI325X $2.00 |
| 4-bit | 0.6 GB | 0.7 GB | 1x MI300X (192 GB) Vultr |
$1.85 | 1x H100 $1.99 · 1x MI325X $2.00 |
Memory is the weights at that precision plus 20% for the runtime and a short context; a long context needs more. Prices are the lowest on-demand hourly rates in the SAVRN Index, read Sep 18, 2026.
Model Card
By LightOn AI, published under apache-2.0, revision 7e3e7b0cb83e.
[!NOTE] LightOnOCR-2 is now available and state-of-the-art on OlmOCR-bench, with new image detection variants! Check it out here: lightonai/LightOnOCR-2-1B
LightOnOCR-1B-1025
Full BF16 version of the model. We recommend this variant for inference and further fine-tuning.
LightOnOCR-1B is a compact, end-to-end vision–language model for Optical Character Recognition (OCR) and document understanding. It achieves state-of-the-art accuracy in its weight class while being several times faster and cheaper than larger general-purpose VLMs.
Paper|Read the full blog post|Try the demo|Finetuning notebook
Highlights
- Speed: 5× faster than dots.ocr, 2× faster than PaddleOCR-VL-0.9B, 1.73× faster than DeepSeekOCR
- Efficiency: Processes 5.71 pages/s on a single H100 (~493k pages/day) for <$0.01 per 1,000 pages
- End-to-End: Fully differentiable, no external OCR pipeline
- Versatile: Handles tables, receipts, forms, multi-column layouts, and math notation
- Compact variants: 32k and 16k vocab options for European languages
Model Overview
Configuration
- Architecture
- LightOnOCRForConditionalGeneration
- Context length (tokens)
- 8,192
- Layers
- 28
- Hidden size
- 1,024
- Feed-forward size
- 3,072
- Attention heads
- 16
- Key/value heads
- 8
- Head dimension
- 128
- Vocabulary size
- 151,936
- RoPE base
- 1,000,000
- Model type
- mistral3
Identity and Version
- Repository
- lightonai/LightOnOCR-1B-1025
- Publisher
- LightOn AI
- Task
- Image to text
- Modality
- Image and text
- Library
- transformers
- Parameters
- 1.2B parameters
- Languages
- en, fr, de, es, it, nl, pt, sv
- Revision
- 7e3e7b0cb83e237e7d237af5a583a002ea632547
- First published
- 2025-10-20
- Last updated
- 2026-06-29
Files and Weights
17 files, 2.3 GB in total. The weights are 1 file totalling 2.3 GB in safetensors.
Weights1 file · 2.3 GB
Configuration7 files · 6.6 KB
Tokenizer4 files · 15.9 MB
Documentation1 file · 7.7 KB
Other3 files · 1.2 MB
Repository1 file · 1.6 KB
Every file
| File | Type | Size | SHA-256 |
|---|---|---|---|
| model.safetensors | Weights | 2.3 GB | c588f8ef5e61 |
| .eval_results/olmocrbench.yaml | Configuration | 2.0 KB | — |
| added_tokens.json | Configuration | 707 B | — |
| config.json | Configuration | 2.1 KB | — |
| generation_config.json | Configuration | 238 B | — |
| preprocessor_config.json | Configuration | 703 B | — |
| processor_config.json | Configuration | 208 B | — |
| special_tokens_map.json | Configuration | 613 B | — |
| README.md | Documentation | 7.7 KB | — |
| chat_template.jinja | Other | 720 B | — |
| lightonocr-banner.png | Other | 1.1 MB | 9c336d908f36 |
| notebook.ipynb | Other | 107.3 KB | — |
| .gitattributes | Repository | 1.6 KB | — |
| merges.txt | Tokenizer | 1.7 MB | — |
| tokenizer.json | Tokenizer | 11.4 MB | aeb13307a71a |
| tokenizer_config.json | Tokenizer | 5.6 KB | — |
| vocab.json | Tokenizer | 2.8 MB | — |
License and Download
- License
- apache-2.0
- Access
- Open weights, no gate
- Download size
- 2.3 GB
Download from LightOn AI
Released by LightOn AI through its official repository on Hugging Face. Read the license.
Built From
- Described by arXiv:2412.13663
- Described by arXiv:2601.14251
Evaluations
Each result is shown as reported, with the conditions its reporter stated. None is a SAVRN measurement. A comparison lines two results up only when their configuration, unit and setup are all stated and identical.
| Benchmark | Conditions | Result | Reported by | Revision | Date |
|---|---|---|---|---|---|
| allenai/olmOCR-bench | Task arxiv_mathMetric arxiv_mathComparison conditions not established | 81.4 | LightOnOCR technical report Reported by a third party |
Evaluated revision not stated | 2026-02-20 |
| allenai/olmOCR-bench | Task baselineMetric baselineComparison conditions not established | 99.6 | LightOnOCR technical report Reported by a third party |
Evaluated revision not stated | 2026-02-20 |
| allenai/olmOCR-bench | Task headers_footersMetric headers_footersSetup Instead of removing headers and footers, our model is trained for full-page transcription and explicitly rewards their presence (via flipped RLVR tests), which lowers this score under the original benchmark scoring.Comparison conditions not established | 35.5 | LightOnOCR technical report Reported by a third party |
Evaluated revision not stated | 2026-02-20 |
| allenai/olmOCR-bench | Task long_tiny_textMetric long_tiny_textComparison conditions not established | 88.7 | LightOnOCR technical report Reported by a third party |
Evaluated revision not stated | 2026-02-20 |
| allenai/olmOCR-bench | Task multi_columnMetric multi_columnComparison conditions not established | 80 | LightOnOCR technical report Reported by a third party |
Evaluated revision not stated | 2026-02-20 |
| allenai/olmOCR-bench | Task old_scansMetric old_scansComparison conditions not established | 35.2 | LightOnOCR technical report Reported by a third party |
Evaluated revision not stated | 2026-02-20 |
| allenai/olmOCR-bench | Task old_scans_mathMetric old_scans_mathComparison conditions not established | 71.6 | LightOnOCR technical report Reported by a third party |
Evaluated revision not stated | 2026-02-20 |
| allenai/olmOCR-bench | Task overallMetric overallSetup Excluding Headers & Footers categoryComparison conditions not established | 76.1 | LightOnOCR technical report Reported by a third party |
Evaluated revision not stated | 2026-02-20 |
| allenai/olmOCR-bench | Task table_testsMetric table_testsComparison conditions not established | 76.4 | LightOnOCR technical report Reported by a third party |
Evaluated revision not stated | 2026-02-20 |
Memory Requirements
| Precision | Weights in memory |
|---|---|
| As published | 2.3 GB |
| 16-bit | 2.3 GB |
| 8-bit | 1.2 GB |
| 4-bit | 0.6 GB |
Weights only, from the published parameter count; the key-value cache and runtime add to this.
Questions About LightOnOCR-1B-1025
How much GPU memory does LightOnOCR-1B-1025 need?
About 2.8 GB at 16-bit and 0.7 GB at 4-bit: the weights (1.2B parameters) plus a working margin. A long context needs more.
What is the cheapest GPU to run LightOnOCR-1B-1025 on?
At 16-bit, 1x MI300X from $1.85 an hour; at 4-bit, 1x MI300X from $1.85 an hour, at the lowest on-demand prices the SAVRN Index lists.
Can I use LightOnOCR-1B-1025 commercially?
Yes. LightOnOCR-1B-1025 is released under Apache License 2.0. The Apache License 2.0 is a permissive open-source license. It permits commercial use, modification and redistribution. It requires keeping the license and copyright notices and any NOTICE file, stating significant changes, and it includes an express patent grant from contributors.
What is LightOnOCR-1B-1025's context length?
8,192 tokens, from the maximum position embeddings in its published configuration.
Similar Models
Join our WeChat and Discord community Use GLM-OCR's API GLM-OCR is a multimodal OCR model for complex document understanding, built on the GLM-V encoder–decoder architecture. It introduces Multi-Token Prediction (MTP) loss and stable full-task reinforcement learning to improve training efficiency, recognition accuracy, and generalization. The model integrates the CogViT visual encoder pre-trained on large-scale image–text data, a lightweight cross-modal connector with efficient token downsampling, and a GLM-0.5B language decoder. Combined with a two-stage pipeline of layout analysis and parallel recognition based on PP-DocLayout-V3, GLM-OCR delivers robust and high-quality OCR performance…
Join our WeChat and Discord community Use GLM-OCR's API GLM-OCR is a multimodal OCR model for complex document understanding, built on the GLM-V encoder–decoder architecture. It introduces Multi-Token Prediction (MTP) loss and stable full-task reinforcement learning to improve training efficiency, recognition accuracy, and generalization. The model integrates the CogViT visual encoder pre-trained on large-scale image–text data, a lightweight cross-modal connector with efficient token downsampling, and a GLM-0.5B language decoder. Combined with a two-stage pipeline of layout analysis and parallel recognition based on PP-DocLayout-V3, GLM-OCR delivers robust and high-quality OCR performance…
This Hub repository contains a HuggingFace's transformers implementation of the original Kosmos-2 model from Microsoft. Use the code below to get started with the model. This model is capable of performing different tasks through changing the prompts. First, let's define a function to run a prompt. Here are the tasks Kosmos-2 could perform: Once you have the entities, you can use the following helper function to draw their bounding bboxes on the image
TrOCR model fine-tuned on the SROIE dataset. It was introduced in the paper TrOCR: Transformer-based Optical Character Recognition with Pre-trained Models by Li et al. and first released in this repository. Disclaimer: The team releasing TrOCR did not write a model card for this model so this model card has been written by the Hugging Face team. The TrOCR model is an encoder-decoder model, consisting of an image Transformer as encoder, and a text Transformer as decoder. The image encoder was initialized from the weights of BEiT, while the text decoder was initialized from the weights of RoBERTa. Images are presented to the model as a sequence of fixed-size patches (resolution 16x16), which…
captioning pretrained on COCO dataset - base architecture (with ViT large backbone). Authors from the paper write in the abstract: Vision-Language Pre-training (VLP) has advanced the performance for many vision-language tasks. However, most existing pre-trained models only excel in either understanding-based tasks or generation-based tasks. Furthermore, performance improvement has been largely achieved by scaling up the dataset with noisy image-text pairs collected from the web, which is a suboptimal source of supervision. In this paper, we propose BLIP, a new VLP framework which transfers flexibly to both vision-language understanding and generation tasks. BLIP effectively utilizes the…