Show HN: OCR Benchmark Focusing on Automation
13 comments
·March 12, 2025kapitalx
Great list! I’ll definitely run your benchmark against Doctly.ai (our PDF-to-Markdown service) specially as we publish our workflow service, to see how we stack up.
One thing I’ve noticed in many benchmarks, though, is the potential for bias. I’m actually working on a post about this issue, so it’s top of mind for me. For example, in the omni benchmark, the ground truth expected a specific order for heading information—like logo, phone number, and customer details. While this data was all located near the top of the document, the exact ordering felt subjective. Should the model prioritize horizontal or vertical scanning? Since the ground truth was created by the company running the benchmark, their model naturally scored the highest for maintaining the same order as the ground-truth.
However, this approach penalized other LLMs for not adhering to the "correct" order, even though the order itself was arguably arbitrary. This kind of bias can skew results and make it harder to evaluate models fairly. I’d love to see benchmarks that account for subjectivity or allow for multiple valid interpretations of document structure.
Did you run into this when looking at the benchmarks?
On a side note, Doctly.ai leverages multiple LLMs to evaluate documents, and runs a tournament with a judge for each page to get the best data (this is only on the Precision Ultra selection).
themanmaran
Hey I wrote the Omni benchmark. I think you might be misreading the methodology on our side. Order on page does not matter in our accuracy scoring. In fact we are only scoring on JSON extraction as a measurement of accuracy. Which is order independent.
We chose this method for all the same reasons you highlight. Text similarity based measurements are very subject to bias, and don't correlate super well with accuracy. I covered the same concepts in the "The case against text-similarity"[1] section of our writeup.
prats226
Bias wrt ordering is a great point. What we consider structured information in this benchmark is irrespective of how its presentation (Order, format etc), it should be directly comparable. So the benchmark does that it into account.
Example is if you are only converting lets say an invoice into markdown, you can introduce bias wrt ordering etc. But if the task is to find out invoice number, total amount, number of line items with headers like price, amount, description, in that case you can compare two outputs without a lot of bias. Eg even if columns are interchanged, you will still get the same metric.
kapitalx
Exactly. You still have to be explicit in order to remove bias. Either by sorting the keys, or looking up specific keys. For arrays, I would say order still matters. For example when you capture a list of invoice items, you should maintain order.
bn-l
In your pricing example I see $6.27 for a 10 page document. That is extremely expensive.
prats226
Can share link? Maybe some kind of mistake.
themanmaran
Love to see another benchmark! We published the OmniAI OCR benchmark the other week. Thanks for adding us to the list.
One question on the "Automation" score in the results, is this a function of extraction accuracy vs the accuracy of the LLM's "confidence score". I noticed the "accuracy" column was very tightly grouped (between 79 & 84%) but the automation score was way more variable.
And side note: is there an open source Mistral benchmark for their latest OCR model? I know they claimed it was 95% accurate, but it looks that was based on an internal evaluation.
prats226
Automation is combination of both, accuracy and accuracy of confidence scores.
Good way to think about automation is recall at high precision which is what you need for true automation where you don't worry about documents that are very likely to have correct results and focus on manually correcting documents likely to have errors.
The reason accuracies are tighly grouped but not the automation is because these models are trained to be accurate but not necessarily predictable, where there is no real way to get confidence score calibrated.
Couldn't find the benchmark mistral used as well
criddell
How do these compare to traditional commercial and open source OCR tools? What about things like the Apple Vision APIs?
sumedh
> Apple Vision APIs
Can we even OCR pdfs using Apple Vision APIs
prats226
Don't think you can. And also there is big difference in plain old OCR, which is just getting all text out from image and document processing which is can you only get the relevant information in a good structure that can be directly pushed into a database.
constantinum
There was a discussion on this benchmark https://getomni.ai/ocr-benchmark couple of weeks ago here > https://news.ycombinator.com/item?id=43118514
prats226
Yeah this new benchmark is kind of inspired by these existing benchmarks, things that are missing wrt automation
OCR/Document extraction field has seen lot of action recently with releases like Mixtral OCR, Andrew Ng's agentic document processing etc. Also there are several benchmarks for OCR, however all testing for something slightly different which make good comparison of models very hard.
To give an example, some models like mixtral-ocr only try to convert a document to markdown format. You have to use another LLM on top of it to get the final result. Some VLM’s directly give structured information like key fields from documents like invoices, but you have to either add business rules on top of it or use some LLM as a judge kind of system to get sense of which output needs to be manually reviewed or can be taken as correct output. No benchmark attempts to measure the actual rate of automation you can achieve.
We have tried to solve this problem with a benchmark that is only applicable for documents/usecases where you are looking for automation and its trying to measure that end to end automation level of different models or systems.
We have collected a dataset that represents documents like invoices etc which are applicable in processes where automation is needed vs are more copilot in nature where you would need to chat with document. Also have annotated these documents and published the dataset and repo so it can be extended.
Here is writeup: https://nanonets.com/automation-benchmark Dataset: https://huggingface.co/datasets/nanonets/nn-auto-bench-ds Github: https://github.com/NanoNets/nn-auto-bench
Looking for suggestions on how this benchmark can be improved further.