Any chance a public API for Cronometer will be available?

2»

Comments

  • Is anyone from Cronometer even reading our comments?

  • I’m wasting time every week downloading CSVs to upload to an AI for analysis. Either add an AI feature or give us an API. Otherwise you are losing customers to other apps. Their nutrition tracking isn’t yet up to Cronometer standards, but that’s going to catch up quickly, and then I will be leaving Cronometer.

  • I'm also shocked this isn't an option yet. I doesn't need to be anything fancy, just let us download raw data as json from a simple API. I also just wanted to use a local AI setup to analyze my nutrition data and combine it with other health markers, but the manual step through CSV is a huge pain.

  • It is counterproductive to not provide an API, guys. The day a similar service will provide it, customers will fly away. Take the courage to do it for those that are paying the annual fee.

  • They obviously have API access because their bloody "User Manual" specifically mentions:
    "Connected Apps: If you have granted an API partner access to your Cronometer account you will see it listed here."
    https://support.cronometer.com/hc/en-us/articles/360018867471-Sharing

    You already have the capability, Chronometer peeps -- now provide your paying users with access!

  • Another voice for an API - this is quickly becoming a minimum requirement for my health app choices.

  • Hi @Karen_Cronometer ; any updates on this?

  • Into the void, but am considering leaving paid sub because no API. Pls add one.

  • Yup, really need an API. Not even asking for an MCP server, just a RESTful API I can call with Claude, which is doing a fantastic job fixing my custom foods using the browser...but that burns so many tokens. HTTP calls would be seriously more efficient, and paying customers have been asking for over half a decade by the looks of it.

  • cphoskins
    edited October 25

    I built an open-source MCP server for Cronometer and have been maintaining it for a while. It works today with Claude Desktop, Claude Code, Cursor and any other MCP-compatible client.

    https://pypi.org/project/cronometer-mcp/
    https://github.com/cphoskins/cronometer-mcp

    What it covers: food log with full macro and micronutrient breakdown, daily summaries, adding and removing diary entries, copying days, recurring foods, macro targets and schedules, fasting history, biometrics, and raw CSV export. Everything runs on your own machine with your own login. Nothing goes through me or any third party.

    Two things to know before you install it:

    • It is unofficial. Cronometer has no public API, so this signs in the same way the web app does. A site change can break it, and you should read the current terms of service and decide for yourself whether you are comfortable with that. It needs a Gold account.
    • I have asked Cronometer for partner API access so this can move onto a supported interface. If anyone there is reading, I'm happy to maintain it under your review or hand it over as the basis for an official connector. Either works.

    Issues and feature requests on GitHub, please. Happy to answer questions here.

  • You can already receive aggregates from Health Connect. Basically, just make an app that reads HC and makes Cronometer sync to HC.

    Cronometer will not provide food macro and calorie information via API because it makes it trivial for competitors to simply read and store the whole food database. Then who will pay for Cronometer?

  • AlexS: In reverse order: I would pay and many others. I have been paying for years and downloading my data via scripts that get it from the website. Ironically it looks like HC would get me approximate timestamps, which is why I started paying for gold.

    Second, cronometer is built on public databases. It is greatly refined and curated. But that's neither here nor there.

    For important stuff: Health Connect (HC) is incomplete. It's not bad, but looking at dinner yesterday I see that it is missing soluble fiber and leucine. It only has some nutrients.

    Next: I actually built software (before LLMs!) to get meal data from Google Fit. Google changed it way too often and then discontinued it. Wasted a lot of my time. It is better practice to give paying customers access to their data via an API instead of punting to an unreliable 3rd party.

    Finally, and most importantly, the API would only give people access to data on food they logged, not the whole database. Easy enough to check for abuse if someone is trying to log all foods.

  • Lack of API is why I won't be renewing. I love Cronometer, but it got infinitely easier to track my food when I was able to use Claude to deduce a recipe for a restaurant meal. Thing is, it had to enter that using the web interface which is slow and super token heavy. An API that I could wrap an MCP server around is what I need to keep using Cronometer.

  • I feel like Chronometer has already missed the boat on this. An API to access our data would have kept Chronometer relevant. Now there's multiple other options in the space that allow full integrations (e.g. Sparkyfitness, Nutritrace, etc.) and nothing really preventing individuals from vibe coding their own solutions.

  • Cronometer licenses data from NCCDB (Nutrition Coordinating Center Food & Nutrient Database), so it cannot simply provide free API access to its entire food database. That data is subject to licensing restrictions.

    Cronometer also has a free plan, so the argument that people need to build their own alternative to avoid paying for it doesn't really hold up.

    Vibe coding isn't much of a threat to Cronometer either. Even if someone builds a competing app using AI, they still need a reliable food database. That means licensing data from sources like Nutritionix or NCCDB, or relying on freely available alternatives.

    And if you can get by using free data from the USDA, why would you even need Cronometer in the first place?

    Cronometer helps you track calories, macros, and micronutrients. But if your goal is simply to track daily calorie intake and monitor your weight, why would you need access to anything beyond aggregate daily totals for calories, protein, carbohydrates, fats, and fiber?

    Which app provides API access to its food database? I can't think of any not MacroFactor, not CalorieCodex, not even MyFitnessPal or Cronometer, if I remember it right.

  • "Cronometer licenses data from NCCDB (Nutrition Coordinating Center Food & Nutrient Database), so it cannot simply provide free API access to its entire food database."

    You're arguing against points nobody made.

    No one is asking for API access to the NCCDB database. We're asking for access to our own data: our logs, our biometrics, and the nutrient totals Cronometer calculates from them. That's user data, not licensed reference data, and plenty of services that license third-party content still give users programmatic access to their own records.

    No one mentioned cost. I built my own solution because I wanted control over my data, not to avoid a subscription.

    Free food databases exist and are solid: USDA FoodData Central, Open Food Facts, the Canadian Nutrient File, the UK's CoFID, and Australia's food composition database, among others.

    On the API question, Nutritionix, FatSecret, and Edamam all offer food-data APIs. But again, that's not what's being requested.

    I'm also not sure why you're defining my goals for me. Some of us track micronutrients and want to analyze them alongside other health and fitness data.

    As for vibe coding: if you don't think it's changing the landscape, or that real alternatives with full integrations already exist, you're not paying attention.

  • We're asking for access to our own data: our logs, our biometrics, and the nutrient totals Cronometer calculates from them

    All of which are already exported to Health Connect ot Health Kit

    I built my own solution because I wanted control over my data, not to avoid a subscription.

    Scraping data when an app purposely does not provide an API is against the TOS.

    I am not sure why you think it's your right to scrape and violate the TOS through vibe coding, yet act entitled to being provided an API to harvest even more data off the app.

    You agreed to the terms before you started using it, and you also paid after seeing that they do not provide API access to the data.

    You don't have the moral high ground here.

    Free food databases exist and are solid: USDA FoodData Central, Open Food Facts, the Canadian Nutrient File, the UK's CoFID, and Australia's food composition database, among others.

    Yet you come to use one with a proprietary database, make it make sense.

    I'm also not sure why you're defining my goals for me. Some of us track micronutrients and want to analyze them alongside other health and fitness data.

    Tracking micros is sort of fantasy; the apple near my store isn't the same apple near you.

    As for vibe coding: if you don't think it's changing the landscape, or that real alternatives with full integrations already exist, you're not paying attention.

    what landscape is changing? Cronometer was a top solution, it's still a top solution to this problem

  • Scraping data when an app purposely does not provide an API is against the TOS.

    I am not sure why you think it's your right to scrape and violate the TOS through vibe coding, yet act entitled to being provided an API to harvest even more data off the app.

    What are you even talking about? No one said anything about scraping data from anywhere. I'm using a custom-built solution that has no interaction with Cronometer in any way. You're obviously just a troll, and I'm done wasting my time here.

  • I'm using a custom-built solution that has no interaction with Cronometer in any way.

    How does such a solution work without breaking TOS? You can enlighten us all.

  • Wow, AlexS is some kind of troll. Dude, we want the data that we are entering from the app we are paying for. Health connect is incomplete. I use python scripts (not vibe coded!) to download the same stuff from the website that anyone can get. This lets me do everything from tracking micronutrients (not in health connect) to aligning food intake with glucose monitoring (Cronometer can't do that as claimed ... it is 3 hours off ... I tried to get them to fix it).

    What are you seeing that breaks the TOS? Be specific!

    Setting aside the fact that I don't feel bound by clickwrap TOS that has changed since I signed up (yeah, it says they can change whatever whenever ... nuts to that) ... they are pretty clear that you can't use their content for LLM training. OK. Yes, automated downloading and traditional analytics arguably fall under that section ... but that is a stretch. I think the thrust is clear: Don't use crono data to train LLMs. Got it. If you want to argue about minor parts of a section really aimed at LLMs, have fun.

    And again: There is no risk that someone will take crono data and do their own app. The foods in my diary will be like 1% of the total. Same for anyone. As long as you are not aggregating across people this is just not an issue for real users.

    And we are missing so much: The two examples I offered. Easily updated recipes. Better food logging. Advanced analytics. This is why I pay for gold. Not CICO. And even on CICO I use the data to periodically refine and improve BMR estimation. There's a whole other thread on that. Crono could do it but don't. Fine. I do it myself, with the data.

    Even garmin, the most closed ecosystem outside of apple, has just launched an API.

    If cronometer were to cut off my access to my data, I'm gone. There is no value for me. For me it is a nice log book paired with a good mapping of foods to nutrients. Great. Value add, I pay. But if they don't let me have the results of the mapping it's no good to me.

  • AlexS
    edited October 27

    And again: There is no risk that someone will take crono data and do their own app. The foods in my diary will be like 1% of the total. Same for anyone. As long as you are not aggregating across people this is just not an issue for real users.

    Calorie-tracking app providers are bound by the TOS of food data providers.

    Food database providers don't let you cache the data for more than 24 hours in many cases.

    Even garmin, the most closed ecosystem outside of apple, has just launched an API.

    Garmin is not even useful for dieting.

    Don't tell you believe the garmin data?

    A 2020 systematic review by Fuller et al. found that consumer wearables often perform poorly at estimating energy expenditure, even when they measure steps and heart rate reasonably well. Across brands like Apple, Garmin, Polar, and Withings, calorie burn estimates frequently showed substantial error, often in the 15–40% range and sometimes exceeding 50%. Wearables can still be useful for tracking trends, but not for precise calorie calculations.

    Most of you guys sound like nerds missing the forest for the trees, tbh.

    Call me a troll or anything, but you guys are just adding up a bunch of incorrect numbers and believing them to produce anything of value.

    Here's a tip: more accurate method is "adaptive tdee" which is way more accurate than apple watch and garmin, lookup data on this.

    Pro tip: use claude/chatgpt for help.

  • AlexS
    edited October 28

    And we are missing so much: The two examples I offered. Easily updated recipes. Better food logging. Advanced analytics. This is why I pay for gold. Not CICO. And even on CICO I use the data to periodically refine and improve BMR estimation. There's a whole other thread on that. Crono could do it but don't. Fine. I do it myself, with the data.

    Before you start using an LLM for "hyper optimization," you need to understand the difference between BMR vs. TDEE.

    How are you using calorie intake data for refining BMR estimation, and how is that useful for dieting?

    TDEE = Total daily energy expenditure, or total calories you burn in a deficit.

    Deficit = TDEE - Intake

    Deficit is directly proportional to your rate of weight loss.

    There's a whole other thread on that. Crono could do it but don't. Fine. I do it myself, with the data.

    Calorie tracking itself is not that "precise", the apple in my garden has different amount of sugar than the apple you eat still they are the same apple in calorie tracking app.

    Packaged foods can have labeling errors of up to 20%, depending on the regulations and jurisdiction.

  • it is incredible that there is STILL no api. I have been wanting to enter this information into my multivariable analysis for a while but not dice...

    @AlexS Go to sleep man. You are not even getting the point of the discussion and are just making "points" to refute arguments no one is making.

  • Such an API is provided by no other apps; it just seems you do not understand the database licensing terms of these apps.

    You should stop blackmailing the devs.