Absolutely loved the article, the process, and the results. Hated the price.
You could pay a human to read receipts, 1 every 30 seconds (that’s slow!), $15/hr (twice the US federal minimum wage!), plus tax and overhead ($15x1.35) comes out to $20.25/hr over 5 hours. $101 all in.
Sure, sure, a human solution doesn’t scale. But this sort of project makes me feel like we haven’t hit the industrialization moment that i thought we had quite yet.
My wife complains about people complaining about the price of eggs every time the subject comes up because it's her duty as a housewife to know about the price of all the protein sources and they are still a bargain -- who'd have thought that the price of transcribing the receipts would be seen as even more onerous?
Receipt scanning OCR has been around for a long time. Circa 2010 I ran enough HITs on Mechanical Turk [1] that I got my own account representative at AWS and I wondered what other kind of HITs other people were running and thought I would "go native" and try making $100 from Turk.
I am pretty good at making judgements for training sets, I have many times made data sets with 2,000-20,000 judgements; I can sustain the 2000 judgements/day of the median Freebase annotator and manage short burst much higher than that with mild perceptual side effects.
I gave up as a Turk though because the other HITs that were easy to find was the task of accurately transcribing cell phone snaps of mangled, damaged, crumpled, torn, poorly printed, poorly photographed or otherwise defective receipts. I can only imagine that these receipts had been rejected by a rather good classical OCR system. The damage was bad enough I could not honestly say I had done a 100% correct job on any single receipt, as I was being asked to do.
[1] in today's lingo: Multimodal with prompts like "Is this a photograph of an X?" and "Write a headline to describe this image"
One issue is that the human was less accurate than the LLM. The other is that the author probably didn't pay $1,500 for this, they probably paid $20 on a subscription.
You're counting just the egg-having receipts, but there were over 11 thousand receipts they had to go through to get to that 500-ish subset. I'm assuming OP wanted to process all of the receipts and then selected just eggs for a simple analytics job. With your rates, the human would cost almost $2000.
Capturing the egg price from known egg receipts was the problem I was focused on, but you're right that there was also a filtering problem in the original spec. You get my upvote for continuing to make the problem interesting for me!
Had the filtering been done during the initial document storage, then the cost would have been much cheaper than your $2,000 estimate. Essentially binning the receipts based on "eggs" or "no eggs" would be free. But, crucially, what happens when the question changes from price per egg to price per gallon of milk? Now the whole stack would need to be sorted again. The $2,000 manual classification would need to be re-applied.
Isn't traditional ML-based classification cheaper for this problem at industrial scale than an LLM though? The OP did of course attempt more traditional generic off-the-shelf OCR tools, but let's consider proper bespoke industrial ML.
Just as a off-the-cuff example, I would probably start with building a tool that locates the date/time from a receipt and takes an image snip of it. Running ONLY image snips through traditional OCR is more successful than trying to extract text from an entire receipt. I would then train a separate tool that extracts images of line items from a receipt that includes item name and price. Yet another tool could then be trained to classify items based on the names of the items purchased, and a final tool to get the price. Now you have price, item, and date to put into your database.
Perhaps generating the training data to train the item classifier is the only place I could see an LLM being more cost effective than a human, but classifying tiny image snips is not the same as one-shotting an entire receipt. As an aside, if there's any desire to discuss how expensive training ML is, don't forget the price to train an LLM as well.
All of this is to say I believe traditional ML is the solution. I'm still not seeing the value prop of LLMs at the industrialization scale outside of very targeted training data generation. A more flippant conclusion might be that we can replace a lot of the parts of data science that makes PhD types get bored with creating traditional ML solutions.
>>So I told Codex “we have unlimited tokens, let’s use them all,” and we pivoted to sending every receipt through Codex for structured extraction. From that one sentence, Codex came back with a parallel worker architecture - sharding, health management, checkpointing, retry logic. The whole thing. When I ran out of tokens on Codex mid-run, it auto-switched to Claude and kept going. I didn’t ask it to do that. I didn’t know it had happened until I read the logs.
----
For anybody still thinking my goodness, how wasteful is this SINGLE EXAMPLE: remember that all of the receipts from the article have helped better-train whichever GPT is deciphering all this thermalprinting.
For a small business owner (like my former self), paying $1500 to have an AI decipher all my receipts is still a heck of a lot cheaper than my accountant's rate. It would also motivate me to actually keep receipts (instead of throw-away/guessing), simply to undaunt the monumental task of recordskeeping.
----
>>But the runs kept crashing. Long CLI jobs died when sessions timed out. The script committed results at end-of-run, so early deaths lost everything. I watched it happen three times. On the fourth attempt I said “I would have expected we start a new process per batch.” That was the fix ... Codex patched it, launched it in a tmux session, and the ETA dropped from 12 hours to 3. Not a hard fix. Just the kind of thing you know after you’ve watched enough overnight jobs die at 3 AM.
>>11,345 receipts processed. The thing that was supposed to take all night finished before I went to bed.
I think at a certain scale we're talking about switching to local trained models which don't have the same operating costs as running a frontier model for OCR. That would reduce the ongoing costs significantly. Might take longer than 30 seconds to read each receipt if you run multiple passes to ensure accuracy, but could run 24/7/365 without the same tax and administration overhead of humans.
Spherical cows aside though, I do agree with you that I should not consider scalability as a given.
I suppose if we had access to a public data set like this receipt bank, programmers could time themselves setting up a solution with off the shelf OCR algos. If they could clock in at under 10 hours they could advertise themselves as being "just as good as an LLM, but significantly cheaper." Downside for the managerial class that wants generative algos for the complete lack of legal protections.
>Everyone needs a rewarding hobby. I’ve been scanning all of my receipts since 2001. I never typed in a single price - just kept the images. I figured someday the technology to read them would catch up, and the data would be interesting.
This is perhaps among the best openers I've ever read.
[spoiler: the tech caught up, the data is interesting]
I usually avoid shallow comments but I feel like this time it has to be said as a conversation starter: That's a lot of eggs!
Also ignoring the benefits of subscriptions, an estimate in the magnitude of thousands of dollars for extracting egg prices still makes me feel like we aren't "there" yet. This should have been a problem with a much more efficient solution given the advancements in the AI, data analysis and OCR space. I am sort of disillusioned.
I actually was going to go for the "why did the chicken not cross the road?". Then I wanted to say "because it was in a price negotiation with the author to sell its eggs", but it was too wordy. Then I thought, "because the author had it as an egg before it could hatch", but it was too dark... Then I gave up.
Well, I guess you cannot make a chicken joke without breaking some eggs (I'll stop now. I'm really sorry, but come on, it's Sunday).
the protagonist is interviewed as a one-man "focus group" in lieu of a national election and one of the questions he is asked is "What do you think about the price of eggs?" and he said roughly "I have no idea, my wife does the shopping."
This is the perfect job for AI, in that it's handling work the human didn't care enough to do manually. Although of course I don't care either. No value judgment there, just an observation. Imagine a place - a field let's say, part of a farm, long ago, but it had a road built through it, and thereby became a non-place, a patch of ground nobody dwells in or pays attention to or cares about, because when they're on it they're always heading somewhere else. The AI phenomenon is like that.
I don't know why people mess with tesseract in 2026, attention-based OCRs (and more recently VLMs) outperformed any LSTM-based approach since at least 2020.
My guess is that it's the entry-point to OCR and the internet is flooded by that, just like pandas for data processing.
The AI writing of the article made me give up halfway through. It’s a neat idea but the writing style of these AI models is brain-grating, especially when it’s the wrong style choice for this kind of technical report.
Inflation adjusted dsta just comes to tell us that either eggs have been outdoing the CPI for 25 years or that actual CPI is way higher than what the BLS calculates.
It depends what dates you're looking at, but energy (gas prices and more) and food (including eggs) are generally recognized as way more volatile than the rest of the CPI.
Eggs were actually quite stable for the 20 years prior to 2001, so maybe don't put your life savings into egg futures...
That is very curious, yes. Eggs seem to just start to increase dramatically after 2000 and indeed outdo the CPI, disregarding the peaks and valleys of the different shocks to egg production like covid and the avian flu.
I read that the price includes free range, eco, etc varieties which are more expensive and in more demand nowadays, probably just that explains a good chunk of the price increase.
Without saying "I bought the exact same brand and type of egg" for 25 years, the data is probably pretty noisy and may reflect the author's income changes as well as the price of eggs.
CPI tracks a weighted average of a large basket of different goods, of which eggs are only a small part. It would be extremely surprising if the change in egg prices over time closely matched CPI.
Apart from the comical cost of extracting this data from paper receipts, is it more likely that stores will publish their product costs over time so trends can be observed or be more like gas stations where no prices are listed. I have no idea why a box of Cheerios costs $7 for processed oats but i see millions of reasons to obscure that data.
I haven't tried it with receipts, but I've gotten excellent OCR results with Gemini 3.0 and now 3.1 on some challenging texts: handwritten letters I couldn't fully decipher myself, vertically printed Japanese texts with tiny furigana readings next to the kanji, a 19th century book in English with extensive use of italics and small caps. Gemini is good at extracting text and formatting from complex layouts, and it might work with egg receipts, too.
Great article through and through. The total number of places you've bought eggs at made me feel a tad depressed though: 4 places where you lived at or spent a longer time, 5 you traveled to *.
I tend to grow bored of a location after a year or two, though I'm certainly in the minority.
* Of course you didn't buy eggs every time you traveled somewhere, so probably not the entire truth.
I think they worked that back from tokens used, hence the estimation, but their actual billing was Claude Code & Codex subscriptions. (Which probably was also the main contributor to it taking 14 days.)
Without 25 years of photographing receipts, weeks of agents coding and billions of token spent, I can predict that egg prices increased, and the graph of my egg consumption over time is concave, part because my income has risen, part because while all prices get inflated, eggs are still cheaper than other sources of protein, and I did in less than 1 microsecond.
I will use them tokens to be able to afford more eggs.
You could pay a human to read receipts, 1 every 30 seconds (that’s slow!), $15/hr (twice the US federal minimum wage!), plus tax and overhead ($15x1.35) comes out to $20.25/hr over 5 hours. $101 all in.
Sure, sure, a human solution doesn’t scale. But this sort of project makes me feel like we haven’t hit the industrialization moment that i thought we had quite yet.
Receipt scanning OCR has been around for a long time. Circa 2010 I ran enough HITs on Mechanical Turk [1] that I got my own account representative at AWS and I wondered what other kind of HITs other people were running and thought I would "go native" and try making $100 from Turk.
I am pretty good at making judgements for training sets, I have many times made data sets with 2,000-20,000 judgements; I can sustain the 2000 judgements/day of the median Freebase annotator and manage short burst much higher than that with mild perceptual side effects.
I gave up as a Turk though because the other HITs that were easy to find was the task of accurately transcribing cell phone snaps of mangled, damaged, crumpled, torn, poorly printed, poorly photographed or otherwise defective receipts. I can only imagine that these receipts had been rejected by a rather good classical OCR system. The damage was bad enough I could not honestly say I had done a 100% correct job on any single receipt, as I was being asked to do.
[1] in today's lingo: Multimodal with prompts like "Is this a photograph of an X?" and "Write a headline to describe this image"
I'd wager gemini flash could get decent results. Id be willing to try on 100 receipts and report cost
Had the filtering been done during the initial document storage, then the cost would have been much cheaper than your $2,000 estimate. Essentially binning the receipts based on "eggs" or "no eggs" would be free. But, crucially, what happens when the question changes from price per egg to price per gallon of milk? Now the whole stack would need to be sorted again. The $2,000 manual classification would need to be re-applied.
Isn't traditional ML-based classification cheaper for this problem at industrial scale than an LLM though? The OP did of course attempt more traditional generic off-the-shelf OCR tools, but let's consider proper bespoke industrial ML.
Just as a off-the-cuff example, I would probably start with building a tool that locates the date/time from a receipt and takes an image snip of it. Running ONLY image snips through traditional OCR is more successful than trying to extract text from an entire receipt. I would then train a separate tool that extracts images of line items from a receipt that includes item name and price. Yet another tool could then be trained to classify items based on the names of the items purchased, and a final tool to get the price. Now you have price, item, and date to put into your database.
Perhaps generating the training data to train the item classifier is the only place I could see an LLM being more cost effective than a human, but classifying tiny image snips is not the same as one-shotting an entire receipt. As an aside, if there's any desire to discuss how expensive training ML is, don't forget the price to train an LLM as well.
All of this is to say I believe traditional ML is the solution. I'm still not seeing the value prop of LLMs at the industrialization scale outside of very targeted training data generation. A more flippant conclusion might be that we can replace a lot of the parts of data science that makes PhD types get bored with creating traditional ML solutions.
>>So I told Codex “we have unlimited tokens, let’s use them all,” and we pivoted to sending every receipt through Codex for structured extraction. From that one sentence, Codex came back with a parallel worker architecture - sharding, health management, checkpointing, retry logic. The whole thing. When I ran out of tokens on Codex mid-run, it auto-switched to Claude and kept going. I didn’t ask it to do that. I didn’t know it had happened until I read the logs.
----
For anybody still thinking my goodness, how wasteful is this SINGLE EXAMPLE: remember that all of the receipts from the article have helped better-train whichever GPT is deciphering all this thermalprinting.
For a small business owner (like my former self), paying $1500 to have an AI decipher all my receipts is still a heck of a lot cheaper than my accountant's rate. It would also motivate me to actually keep receipts (instead of throw-away/guessing), simply to undaunt the monumental task of recordskeeping.
----
>>But the runs kept crashing. Long CLI jobs died when sessions timed out. The script committed results at end-of-run, so early deaths lost everything. I watched it happen three times. On the fourth attempt I said “I would have expected we start a new process per batch.” That was the fix ... Codex patched it, launched it in a tmux session, and the ETA dropped from 12 hours to 3. Not a hard fix. Just the kind of thing you know after you’ve watched enough overnight jobs die at 3 AM.
>>11,345 receipts processed. The thing that was supposed to take all night finished before I went to bed.
Spherical cows aside though, I do agree with you that I should not consider scalability as a given.
This is perhaps among the best openers I've ever read.
[spoiler: the tech caught up, the data is interesting]
I read a lot. This article, entirely.
Also ignoring the benefits of subscriptions, an estimate in the magnitude of thousands of dollars for extracting egg prices still makes me feel like we aren't "there" yet. This should have been a problem with a much more efficient solution given the advancements in the AI, data analysis and OCR space. I am sort of disillusioned.
There's got to be a "it's a chicken/egg problem" joke in there somewhere, but i'm not seeing it.
Well, I guess you cannot make a chicken joke without breaking some eggs (I'll stop now. I'm really sorry, but come on, it's Sunday).
FWIW, you made an eggceptional attempt :).
Less than one per day, assuming they're doing groceries only for themselves
https://en.wikipedia.org/wiki/Franchise_(short_story)
the protagonist is interviewed as a one-man "focus group" in lieu of a national election and one of the questions he is asked is "What do you think about the price of eggs?" and he said roughly "I have no idea, my wife does the shopping."
My guess is that it's the entry-point to OCR and the internet is flooded by that, just like pandas for data processing.
Leaving a comment so I can more easily find this
And for the people wondering about Pandas, use Polars instead
Eggs were actually quite stable for the 20 years prior to 2001, so maybe don't put your life savings into egg futures...
Egg prices: https://fred.stlouisfed.org/series/APU0000708111
CPI: https://fred.stlouisfed.org/series/CPIAUCSL
Core CPI (without food + energy prices): https://fred.stlouisfed.org/series/CPILFESL
I read that the price includes free range, eco, etc varieties which are more expensive and in more demand nowadays, probably just that explains a good chunk of the price increase.
https://www.thebignewsletter.com/p/hatching-a-conspiracy-a-b...
...
>>These are the days of miracle and wonder. I can’t wait to see what [the next] 30 years of eggs looks like.
I tend to grow bored of a location after a year or two, though I'm certainly in the minority.
* Of course you didn't buy eggs every time you traveled somewhere, so probably not the entire truth.
I can assume this person does in fact NOT need to worry about the price of eggs ?
...
> I can’t wait to see what 30 years of eggs looks like.
At $2.70 per receipt, i'd be in no hurry to find out!
I will use them tokens to be able to afford more eggs.