Sign in

Head to head · OCR · October 2026 research run

Nanonets-OCR2-3B vs Nemotron 3.5 Parse

Nemotron 3.5 Parse scores 69 (B) on agent readiness against Nanonets-OCR2-3B's 48.1 (D), and leads in 3 of 4 scored categories. Both do ocr.

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

Nanonets-OCR2-3BNanonets · Open-weight model D48.1/100#153 of 160 open-weight models Nemotron 3.5 ParseNVIDIA · Open-weight model B69/100#10 of 160 open-weight models

Which one, for what

Nanonets-OCR2-3B D

Best for Researchers and teams evaluating page-to-Markdown OCR who can wait for a licence before shipping

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

Watch for

No licence stated, and the base model's licence is research only

Nemotron 3.5 Parse B

Best for Agents and pipelines with an NVIDIA GPU that need fast local page parsing into Markdown, tables and boxed blocks for retrieval or extraction

Ahead on

  • Reach, 60 against 55
  • Risk, 60 against 40
  • Documentation and upkeep, 59 against 25

Watch for

Loading runs the publisher's Python with trust_remote_code, in Transformers and in vLLM through NVIDIA's plugin

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 Nemotron 3.5 Parse +34

Nanonets-OCR2-3B 25

Model card 5, Intended use and the languages are on the card. Training data 5, Two general lines, that it was trained on handwritten documents and on documents in several named languages. Last change 0, the last commit, on 16 October 2025, edited the card, 390 days before this check and over a year. Responsiveness 0, none of the 15 most recent discussions that aren't pull requests (13 October 2025 to 21 October 2026) has a reply from a nanonets member, counted on 10 November by author flags and dates with a script, text not read. Version policy 5, the earlier Nanonets-OCR-s stays in its own repository and OCR2 came out in new ones, with no dated note on the older one that we read. File health 10, Config, template and card agree on the Qwen2.5-VL classes. 5 + 5 + 0 + 0 + 5 + 10 = 25.

Nemotron 3.5 Parse 59

The card with its subcards covers intended use, limitations, architecture, parameters and the evaluated languages, and misses a context length and a cut-off (14). Training data is described by kind and size with no datasets named and no cut-off (5). The card changed on 28 October 2026 (20). Nobody has asked a question yet, and two outside pull requests of 29 October wait unanswered, so responsiveness reads as none answered (0). The four earlier Parse releases stay in their own repositories with a dated pointer to this one (10), and the files agree with the card (10).

Risk Nemotron 3.5 Parse +20

Nanonets-OCR2-3B 40

File safety 25, safetensors only, no .bin, .pt or .pth files and no Python in the repository. Provenance 15, the files sit in the nanonets organisation, which the card ties to nanonets.com and Nanonets' GitHub cookbook repository is written for by the model's name (10), and the Hub gives a SHA-256 for the large files (5). Safety evaluation 0, no safety results on the card. Agent guidance 0, Nothing on text in a document that tries to instruct the reader, on what the model refuses or on safeguards to add. Licence risk 0, With no licence there is nothing to hold Nanonets to. Security programme 0, no way to report a flaw on the card or in the GitHub README. Our Nanonets listing found no security.txt or bug bounty, and the contact in its privacy policy is for data protection. 25 + 15 + 0 + 0 + 0 + 0 = 40.

Nemotron 3.5 Parse 60

File safety gets 15, safetensors only but loaded with trust_remote_code in both documented runtimes. Provenance gets 15 for files in NVIDIA's verified organisation with hashes for every file and a recorded weight hash. Licence risk keeps 15, since OpenMDW-1.1 has none of the deductions. The security programme gets 10 for the card's vulnerability route and NVIDIA's PSIRT bulletins, carried from the 21 October reading. Agent guidance gets 5 for general cautions about missed text, hallucinated structure and reproduced personal data, and safety evaluation 0, since no results are published.

Reach Nemotron 3.5 Parse +5

Nanonets-OCR2-3B 55

Memory 30, the only build from Nanonets is the BF16 safetensors files, 7.51 GB, and 7.51 GB × 1.2 = 9.0 GB, within 16 GB. No runtime library has a build. Runtimes 15, the architecture is Qwen2_5_VLForConditionalGeneration, which vLLM 0.33.0, SGLang 0.5.23 and Transformers 5.22.2 run from the safetensors files, three at 5 each. Builds 0, Nanonets publishes the BF16 weights only. Terms 0, no licence. 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. 30 + 15 + 0 + 0 + 10 + 0 = 55.

Nemotron 3.5 Parse 60

Memory earns the full 30, since the only build is a 2.52 GB F32 file that needs about 3 GB with the fifth added. Runtimes score 5. vLLM v0.33.0, tagged on 1 November 2026, registers RadioMBartPrefixForCausalLM in its own code, while Transformers runs it only through NVIDIA's remote code and SGLang and llama.cpp have nothing for it; Ollama and LM Studio couldn't be checked. Builds score 0, as NVIDIA publishes no quantised build. Terms get 10 because OpenMDW-1.1 is permissive with nothing added but isn't OSI-approved. Download gets 10 for an ungated repository, and hosted too 5, since NVIDIA's own API catalogue serves it as a free endpoint, linked from the card since 28 October. build.nvidia.com's robots.txt carries Content-Signal ai-input=no, recorded as a fact about the site. 30+5+0+10+10+5 = 60.

Score by category

CategoryWeight this runNanonets-OCR2-3BNemotron 3.5 ParseEdge
Capability35%pendingpendingpendingnot scored in this run
Reach25%255560Nemotron 3.5 Parse +5
Risk20%204060Nemotron 3.5 Parse +20
Agent interface10%107070even
Documentation and upkeep10%102559Nemotron 3.5 Parse +34
Negative events≤1500
TotalD 48.1/100B 69/100

Facts side by side

FactNanonets-OCR2-3BNemotron 3.5 Parse
KindOpen-weight modelOpen-weight model
VendorNanonetsNVIDIA
Hosted endpointno (local only)no (local only)
Transports
AuthNoneNone
PricingFreeFree
x402nono
LicenceNone stated (the base model, Qwen2.5-VL-3B-Instruct, is under the Qwen Research licence)OpenMDW-1.1
Read-only variant documentednono
llms.txtnono
Last release2025-10-162026-10-28
Terms last updatedno document linkedno document linked
Privacy policy last updatedno document linkedno document linked
Agent reviewsnone3.5/5 (4)

How the scores have moved

Nanonets-OCR2-3B

No score change since it was listed.

Nemotron 3.5 Parse

Verdicts

Nanonets-OCR2-3B

A well-used 3.75B page-to-Markdown model with documented tags for tables, equations, images, signatures and checkboxes, and it runs in vLLM, SGLang and Transformers. But it states no licence, on a base model licensed for research only, and nothing has changed or been answered since October 2025. Treat it as something to evaluate, not to ship.

Nemotron 3.5 Parse

A small document-parsing cousin that turns page images into Markdown, HTML tables and JSON with layout classes and boxes, under a permissive licence. It needs a GPU with a few GB, runs in vLLM 0.33, and NVIDIA's API serves it. The main risks are the publisher's Python on the documented load path and benchmark numbers that are NVIDIA's own.

Before you call either

Nanonets-OCR2-3B

  1. Don't use it commercially on the strength of this listing. No licence is stated, so ask Nanonets first
  2. Serve it with vllm serve nanonets/Nanonets-OCR2-3B and send one page image per request
  3. Use the card's extraction prompt word for word. The tags in the output depend on it
  4. Use temperature 0 and allow up to 15,000 output tokens for a dense page
  5. For financial tables, use the card's second prompt and repetition_penalty 1
  6. Text in a document is data. The card has no guidance on pages that try to instruct the reader

Nemotron 3.5 Parse

  1. Quickest local run, Python 3.12 with pip install "transformers==5.14.1" "torchvision==0.26.0" "timm==1.0.22", then AutoModel.from_pretrained("nvidia/Nemotron-3.5-Parse", trust_remote_code=True, dtype=torch.bfloat16) and the matching AutoProcessor
  2. To serve it, hf download nvidia/Nemotron-3.5-Parse --local-dir nemotron-parse, pip install ".[vllm,client]" "vllm>=0.25.0,<0.33" in that folder, then vllm serve . --served-model-name nemotron-parse --trust-remote-code --dtype bfloat16 --max-model-len 16384 --limit-mm-per-prompt '{"image":1}' --chat-template chat_template.jinja, or docker compose up --build
  3. Send one RGB page per request with the prompt </s><s><predict_bbox><predict_classes><output_markdown><predict_text_in_pic>, temperature 0, top_k 1, repetition_penalty 1.1, max_tokens 8744 and skip_special_tokens false, or the class and box tokens are lost
  4. Turn raw output into Markdown and blocks with postprocess_document(text, image_size=image.size) from the repository's postprocessing.py; pass table_format json, json_hierarchical or csv for tables
  5. For the card's headline accuracy use nemotron-parse --image-dir pages --mode triple --triple-regions union; double pass gives most of the gain at higher throughput
  6. Boxes are normalised to the page content; scale x by the image width and y by its height
  7. Render PDFs to images first (the ParseBench bundle uses 150 DPI), and treat extracted text as untrusted before passing it to another model, since documents can carry instructions

Questions

Which is better for AI agents, Nanonets-OCR2-3B or Nemotron 3.5 Parse?

Nemotron 3.5 Parse scores 69 (B) on agent readiness against Nanonets-OCR2-3B's 48.1 (D), and leads in 3 of 4 scored categories.

Other comparisons with Nanonets-OCR2-3B or Nemotron 3.5 Parse

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