Eating out is one of life's genuine pleasures. But if you are tracking your nutrition, a restaurant meal can feel like stepping into a fog. There is no barcode to scan, no packaging with a neat nutrition label, and no way of knowing exactly how much oil the chef used. The result? Most people either skip logging entirely or take a wild guess that is usually 30-50% off. Neither approach is particularly helpful.
Tracking food at home is relatively straightforward. You can weigh ingredients, scan barcodes, and check packaging. Restaurants strip away every one of those tools.
Unlike supermarket food, restaurant dishes have no barcode and no standardised nutrition panel. Even when a chain publishes calorie data, the actual plate in front of you may vary from the official figure depending on who made it and how generous they were with the sauce.
Restaurants want food to taste exceptional, which means liberal use of butter, oil, cream, and sugar. A simple-looking side of vegetables might be tossed in 20-30ml of olive oil (180-270 kcal) that you would never use at home. Research published by the BMJ found the average UK restaurant main course contains around 1,033 calories — roughly double a typical home-cooked meal.
A chicken breast at home might be 150g. The same dish at a restaurant could be 250g or more. When you cannot weigh anything, portion estimation becomes the single biggest source of error in your food diary.
AI meal scanning uses your phone camera and machine learning to identify the foods on your plate. Here is the typical process:
NutraSafe's AI meal scanner uses this approach and is designed specifically for UK meals. That means it recognises dishes like a Sunday roast, fish and chips, or a Wagamama katsu curry, not just American-style portion sizes.
AI scanning is typically accurate to within 15-25% for calories. That sounds imprecise, but most people underestimate restaurant meals by 30-50% when guessing unaided. Using AI is a significant improvement over gut feeling.
Since April 2022, businesses in England with 250 or more employees have been required to display calorie information on menus. This covers the majority of UK chain restaurants. Here are some approximate figures for popular dishes to give you a sense of scale:
| Restaurant | Dish | Approx. Calories |
|---|---|---|
| Nando's | Half chicken, medium spice | 550 kcal |
| Wagamama | Chicken katsu curry | 1,100 kcal |
| Pizza Express | Margherita Romana | 760 kcal |
| Greggs | Sausage roll | 327 kcal |
| McDonald's | Big Mac meal (medium) | 845 kcal |
| Five Guys | Cheeseburger | 840 kcal |
| Dishoom | House black daal | 530 kcal |
These figures come from the restaurants' own published data. Bear in mind that independent restaurants and smaller chains are not required to publish calorie information, which is where AI scanning becomes especially valuable.
Even with AI scanning, having a few mental shortcuts for portion estimation is genuinely useful:
Sauces are one of the biggest hidden calorie sources in restaurant meals. A creamy pasta sauce can add 200-400 kcal that is almost impossible to estimate when it is mixed in. Asking for it on the side lets you control the amount and makes AI scanning more accurate too.
Grilled, steamed, or baked dishes tend to be lower in added fats than fried, sauteed, or roasted options. If you are trying to keep your meal within a certain calorie range, the cooking method matters as much as the ingredients themselves.
One of the biggest barriers to food tracking is the social awkwardness of it. Nobody wants to be the person at the table photographing their food for five minutes while everyone else starts eating. Here is how to keep things natural:
When your food arrives, take one quick photo before you pick up your fork. It takes about two seconds. Everyone takes food photos these days — nobody will think twice about it. Then put your phone away and enjoy the meal.
There is no rule that says you must log food the moment you eat it. Many people find it easier to log their restaurant meal when they get home, using the photo they took as a reference. NutraSafe lets you add meals retroactively, so you can review your photo in your own time and get a more considered estimate.
A restaurant meal logged at 80% accuracy is infinitely more useful than one not logged at all. Tracking is about building awareness over time, not about hitting exact numbers on every single meal. If you eat out once or twice a week, a reasonable estimate keeps your weekly average meaningful without turning dinner into a maths exercise.
If you know where you are eating in advance, check the restaurant's menu online beforehand. Most chain restaurants publish full nutrition data on their websites. You can pre-log your meal before you even arrive, then adjust if you end up ordering something different.
There are two schools of thought here, and both are valid:
Some people prefer to log immediately because the details are fresh. If you ordered a starter, main, side, and dessert, you might forget the starter by the time you get home. The downside is that it can interrupt the social flow and feel overly obsessive, especially if you are dining with people who do not track their food.
This is the approach most nutritionists and dietitians recommend. Take a photo when your food arrives, enjoy the experience, and log everything when you get home. If you are worried about forgetting dishes, jot a quick note in your phone between courses. The photo-plus-note approach captures 95% of what you need without any social awkwardness.
The NHS advises that the most effective food diary is one you actually maintain consistently. If logging during meals makes you less likely to eat out or enjoy your food, it is counterproductive. Find the approach that works for your lifestyle.
NutraSafe's AI meal scanner identifies UK dishes, estimates portions, and logs your meal from a single photo. Snap, enjoy, log later.
Download NutraSafe FreeYes. AI-powered meal scanning apps like NutraSafe let you take a photo of your plate and get an estimated calorie and macro breakdown. The AI identifies individual foods, estimates portion sizes, and pulls nutritional data. It is not as precise as weighing ingredients, but it gives a solid ballpark figure that is far better than not logging at all.
AI meal scanning is typically accurate to within 15-25% for calorie estimates, depending on the complexity of the dish. Simple plates like grilled chicken with vegetables are more accurate than heavily sauced or layered dishes. For context, most people underestimate restaurant calories by 30-50% when guessing, so AI scanning is a significant improvement.
Either approach works, but logging after the meal is generally better for your social experience and mental health. Take a quick photo when your food arrives, then put your phone away. You can log the details later at home. The photo gives you an accurate reference so you do not have to rely on memory alone.
Since April 2022, large businesses in England with 250 or more employees are required to display calorie information on menus, including restaurants, cafes, and takeaways. This applies to dine-in food and covers chains like Nando's, Wagamama, Pizza Express, and most fast food outlets. Smaller independent restaurants are not required to provide this information.
Research published by the BMJ found that restaurant meals in the UK contain an average of 1,033 calories per main course, compared to around 500-600 for a typical home-cooked meal. Restaurants use more oil, butter, and larger portions than most people would at home. This does not mean you should avoid eating out, but being aware of the difference helps you plan your day accordingly.
Last updated: February 2026