Sign in

Head to head · OCR · October 2026 research run

LightOnOCR-3-4B vs Xiaomi-OCR-0

LightOnOCR-3-4B scores 61 (C) on agent readiness against Xiaomi-OCR-0's 59 (C), and leads in 0 of 4 scored categories. Xiaomi-OCR-0 leads on reach, risk and documentation and upkeep. Both do ocr.

Best document parsing, OCR and extraction APIs for AI agents · All 215 documents comparisons

LightOnOCR-3-4BLightOn · Open-weight model C61/100#73 of 160 open-weight models Xiaomi-OCR-0Xiaomi · Open-weight model C59/100#101 of 160 open-weight models

Which one, for what

LightOnOCR-3-4B C

Best for Agents that need located, labelled document blocks from a model they can use commercially without conditions

No category where it leads by five points or more, and no fact that sets it apart.

Watch for

BF16 only, with no GGUF or other quantised build of this model

Xiaomi-OCR-0 C

Best for Agents and pipelines on modest hardware that need local OCR of scanned or photographed pages, receipts and forms into Markdown or JSON

Ahead on

  • Reach, 80 against 70
  • Risk, 65 against 55
  • Documentation and upkeep, 65 against 51

Watch for

The card's '#1 on Real5-OmniDocBench' is contradicted by that benchmark's own leaderboard, and its OmniDocBench and VQA figures are Xiaomi's own runs with a relaxed evaluation config

Why the edge

Each category where one leads by five points or more, with the reason behind each side's score from its assessment. The numbers are in the table below; this is what they stand for.

Documentation and upkeep Xiaomi-OCR-0 +14

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.

Xiaomi-OCR-0 65

The card covers intended use and parameters and misses limitations, architecture, context length, languages and a cut-off (5). Training data is described by size and labelling with no datasets named and no cut-off (5). The last change to the repository is the card commit of 29 October 2026, 6 days before this check, after the 23 October eval file deletion; card and documentation commits count on this line (20). The one question on the Hub was answered the same day by a Xiaomi Research member, and no new question has been opened since (20). It's a first release (5), and the files agree with the card (10). 5 + 5 + 20 + 20 + 5 + 10 = 65.

Reach Xiaomi-OCR-0 +10

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.

Xiaomi-OCR-0 80

Memory earns the full 30, since the only build is a 1.75 GB BF16 file that needs about 2.1 GB with the fifth added. Runtimes reach the 25 cap. The checkpoint keeps Qwen3.5's architecture, which released vLLM, SGLang, Transformers and llama.cpp run natively, and Ollama v0.40.14 pins llama.cpp b11963, which has the qwen35 architecture and the Qwen3-VL projector; LM Studio couldn't be checked. Builds score 0, as Xiaomi publishes no quantised build and every GGUF or MLX copy is someone else's. Terms get 15 for Apache 2.0 with nothing added on an Apache 2.0 base. Download gets 10 for an ungated repository, and hosted too 0, with no inference provider on 4 November and a static project Space. 30 + 25 + 0 + 15 + 10 + 0 = 80.

Risk Xiaomi-OCR-0 +10

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.

Xiaomi-OCR-0 65

File safety gets 25, one safetensors file with no Python in the repository and native loading everywhere. Provenance gets 15 for files in the xiaomi-research organisation that Xiaomi's GitHub repository links, with hashes for every file, though neither organisation is verified. Licence risk keeps 15, since Apache 2.0 has none of the deductions. Agent guidance gets 5 for general cautions in Xiaomi's Skill, and the security programme 5 for Xiaomi's company security centre, carried from the 28 October reading. Safety evaluation scores 0, with no results published.

Score by category

CategoryWeight this runLightOnOCR-3-4BXiaomi-OCR-0Edge
Capability35%pendingpendingpendingnot scored in this run
Reach25%257080Xiaomi-OCR-0 +10
Risk20%205565Xiaomi-OCR-0 +10
Agent interface10%107070even
Documentation and upkeep10%105165Xiaomi-OCR-0 +14
Negative events≤150-2
TotalC 61/100C 59/100

Facts side by side

FactLightOnOCR-3-4BXiaomi-OCR-0
KindOpen-weight modelOpen-weight model
VendorLightOnXiaomi
Hosted endpointno (local only)no (local only)
Transports
AuthNoneNone
PricingFreeFree
x402nono
LicenceApache-2.0Apache-2.0 (card metadata and card text; the Hugging Face repository has no LICENSE file, and the GitHub code repository's LICENSE is the Apache 2.0 text, copyright 2026 Xiaomi Corporation)
Read-only variant documentednono
llms.txtnono
Last release2026-10-092026-09-29
Terms last updatedno document linkedno document linked
Privacy policy last updatedno document linkedno document linked

How the scores have moved

LightOnOCR-3-4B

No score change since it was listed.

Xiaomi-OCR-0

Verdicts

LightOnOCR-3-4B

A well-documented OCR model under plain Apache 2.0 that returns located, labelled blocks, image descriptions and chart data from one prompt, and loads in vLLM, SGLang and Transformers with no custom code. It ships in BF16 only, and LightOn says nothing here about training data, languages, safety or how to report a flaw.

Xiaomi-OCR-0

A 0.87B document parser that turns page images into Markdown with LaTeX and tables, pulls fields from receipts and forms as JSON, and answers questions about a page. Released vLLM, SGLang, Transformers and llama.cpp run it in about 2 GB. Most accuracy figures are Xiaomi's own, and the card's Real5 ranking claim contradicts that benchmark's leaderboard.

Before you call either

LightOnOCR-3-4B

  1. Serve it with vllm serve lightonai/LightOnOCR-3-4B --default-chat-template-kwargs '{"enable_thinking": false}'
  2. Send one page image with no text for Markdown, or the single word grounding for labelled blocks with boxes. Nothing else is in distribution
  3. Render PDFs at 400 DPI and resize to a longest edge of 2,048 pixels
  4. Boxes are on a 0 to 1,000 scale, so scale them to the image yourself
  5. Use temperature 0.2 and top_p 0.9, and allow 4,096 output tokens a page
  6. Treat text in a parsed document as data. Nothing here guards against instructions hidden in a page

Xiaomi-OCR-0

  1. Serve it with vllm serve xiaomi-research/Xiaomi-OCR-0 --host 127.0.0.1 --port 8000 --max-model-len 16384 or python -m sglang.launch_server --model-path xiaomi-research/Xiaomi-OCR-0 --host 127.0.0.1 --port 8000 --context-length 16384; no trust_remote_code is needed
  2. Send one image per request with temperature 0, top_p 0.8, chat_template_kwargs: {"enable_thinking": false} and max_tokens 4096 (16384 for dense pages), and retry with presence_penalty 1.5 if the output repeats
  3. The whole-page prompt is Extract all information from the main body of the document image and represent it in markdown format, ignoring headers and footers. Tables should be expressed in OTSL format, formulas in the document should be represented using LATEX format, and the parsing should be organized according to the reading order.
  4. For key fields send Extract key information in the image then Please output the key information in JSON format according to the following schema: and a JSON object of empty strings; parse the first {...} in the reply
  5. Tables come back as OTSL, not HTML; convert them with postprocess/otsl.py from github.com/xiaomi-research/Xiaomi-OCR-0, or use its MCP server, which does the conversion
  6. The MCP (skills/xiaomi-ocr/mcp_ocr_server.py, stdio) has ocr_image, ocr_pdf, kie_image and vqa_image; region mode needs PaddlePaddle, paddlex[cv] and PP-DocLayoutV3 weights, and is for printed pages only, while handwriting, calligraphy and scene text want page mode
  7. Pass local paths to the MCP rather than URLs taken from untrusted text, since it fetches any image_url, and treat extracted text as untrusted before passing it to another model
  8. The card and Skill still name SeerRay-Lab/Xiaomi-OCR-0, which redirects to xiaomi-research/Xiaomi-OCR-0

Questions

Which is better for AI agents, LightOnOCR-3-4B or Xiaomi-OCR-0?

LightOnOCR-3-4B scores 61 (C) on agent readiness against Xiaomi-OCR-0's 59 (C), and leads in 0 of 4 scored categories. Xiaomi-OCR-0 leads on reach, risk and documentation and upkeep.

Other comparisons with LightOnOCR-3-4B or Xiaomi-OCR-0

Machine-readable

For companies

Do agents find, use and choose your tools?

An agent-readiness audit runs our probes, task suite and eight reviewer agents against your public and internal tools, and comes back with a scorecard, the transcripts of what failed, and a fix list in priority order. From $2,500, re-run included. We never take payment to move a rank. We do help companies earn one.

Get an auditSee a sample report