DeepSeek-OCR 2 vs LightOnOCR-3-4B

DeepSeek-OCR 2 35

Chat template 10, no Jinja template in the repository and none in tokenizer_config.json. Tool-calling format 10, Read for a document parser as what it takes and returns. Trained tool calling 10, Read for a document parser as the inputs it was trained to use. Structured output 5, Read for a document parser as typed outputs. Reasoning control 0, A document parser with no reasoning. Context 0, config.json gives 8,192 positions, below the 32K band. 10 + 10 + 10 + 5 + 0 + 0 = 35.

LightOnOCR-3-4B 70

Chat template 20, chat_template.jinja in the repository (7,759 characters), Qwen 3.5's template with thinking off by default. Tool-calling format 25, Read for a document parser as what it takes and returns. Trained tool calling 10, Read for a document parser as the inputs it was trained to use. Structured output 15, Read for a document parser as typed outputs. Reasoning control 0, the card says it was trained and evaluated with thinking off, the template's default, and documents no thinking mode. Context 0, LightOn states no context length for it. 20 + 25 + 10 + 15 + 0 + 0 = 70.

DeepSeek-OCR 2 22

Model card 2, Intended use is there, as the two prompts. Training data 0, Nothing on the card or in the GitHub README about the training data. Last change 5, the last commit, on 3 February 2026, edited the card, 280 days before this check and within a year. Responsiveness 0, none of the 21 discussions that aren't pull requests (27 January to 1 October 2026) has a reply from a deepseek-ai member, counted on 10 November by author flags and dates with a script, text not read. Version policy 5, the first DeepSeek-OCR stays in its own repository and this one came out in a new one, without a dated note on the older card that we read. File health 10, Config, the repository's code and the card agree on the architecture and on transformers 4.46.3. 2 + 0 + 5 + 0 + 5 + 10 = 22.

LightOnOCR-3-4B 51

Model card 11, Intended use, a limitation (instructions other than its two prompts are out of distribution), the architecture (Qwen 3.5's) and the size (4B) are on the card. Training data 0, Nothing on the card about the training data. Last change 15, one commit, the release on 9 October 2026, 32 days before this check and within 90. Responsiveness 10, Of the 2 discussions that aren't pull requests (both opened on 9 October 2026), 1 has a reply from a lightonai member, which is some. Version policy 5, LightOnOCR-2-1B stays in its own repository, and a commit to its card on 8 October 2026 named this model as its new version. File health 10, Config, template and card agree on the Qwen 3.5 classes and on thinking off by default. 11 + 0 + 15 + 10 + 5 + 10 = 51.

DeepSeek-OCR 2 65

Memory 30, the only build from DeepSeek is the BF16 safetensors file, 6.78 GB, and 6.78 GB × 1.2 = 8.1 GB, within 16 GB. No runtime library has a build of this version. Runtimes 10, Released vLLM 0.33.0 registers DeepseekOCR2ForCausalLM (deepseek_ocr2.py), and llama.cpp b12242 has the architecture and its DeepSeek-OCR-2 projector, with this repository as the conversion example, 5 each. Builds 0, DeepSeek publishes the BF16 weights only. Terms 15, Apache 2.0, OSI-approved, with nothing added, in the card's metadata and as LICENSE.txt in the repository. Download 10, the Hub API reports gated false on 10 November 2026, so the files download with no account and nothing to accept. Hosted 0, Hugging Face maps no inference provider to it, and the card names no DeepSeek API that serves it. 30 + 10 + 0 + 15 + 10 + 0 = 65.

LightOnOCR-3-4B 70

Memory 30, the only build from LightOn is the BF16 safetensors file, 9.08 GB, and 9.08 GB × 1.2 = 10.9 GB, within 16 GB. No runtime library has a build of this model. Runtimes 15, the architecture is Qwen3_5ForConditionalGeneration, which vLLM 0.33.0, SGLang 0.5.23 and Transformers 5.22.2 run from the safetensors file, three at 5 each. Builds 0, LightOn publishes the BF16 weights only. Terms 15, Apache 2.0, OSI-approved, with nothing added, in the card's metadata and licence section. Download 10, the Hub API reports gated false on 10 November 2026, so the files download with no account and nothing to accept. Hosted 0, Hugging Face maps no inference provider to it, and the card names no hosted API for it. 30 + 15 + 0 + 15 + 10 + 0 = 70.

DeepSeek-OCR 2 50

File safety 15, safetensors only, no .bin, .pt or .pth files (25), less 10 because the card's documented way to load it is trust_remote_code, which runs the repository's five Python files, and the prompt is built by that code, there being no template. Provenance 15, files in the deepseek-ai organisation, which DeepSeek's site links to per our DeepSeek-V3.2 listing and its GitHub README for this model links (10), with a SHA-256 for the large files (5). Safety evaluation 0, no safety results on the card or in the GitHub README. Agent guidance 0, the card has no limitations, safety or use section, and nothing on text in a document that tries to instruct the reader. Licence risk 15, Apache 2.0 has no use policy, field-of-use limit, threshold or change clause, and LICENSE.txt is in the repository. Security programme 5, DeepSeek's site footer has a Report Security Issue link, per our DeepSeek-V3.2 listing (5). 15 + 15 + 0 + 0 + 15 + 5 = 50.

LightOnOCR-3-4B 55

File safety 25, one safetensors file, no .bin, .pt or .pth files and no Python in the repository. Provenance 15, the files sit in the lightonai organisation, which the Hub marks as verified and LightOn's GitHub repository for the models links to (10), and the Hub gives a SHA-256 for the large files (5). Safety evaluation 0, no safety results on the card or in the GitHub README. Agent guidance 0, Nothing on text in a document that tries to instruct the reader, or on safeguards to add. Licence risk 15, Apache 2.0 has no use policy, field-of-use limit, threshold or change clause, and nothing is added on the card. Security programme 0, no way to report a flaw found on the card or in the GitHub README. lighton.ai couldn't be read for a security.txt. 25 + 15 + 0 + 0 + 15 + 0 = 55.