Say what you ate out loud — "two hundred grams of cottage cheese with honey" — and this bot transcribes it and logs KBZU automatically, with photo and text as fallback inputs and a minimal daily-summary interface.
NutriTrackerAI_Bot (branded Ёжик-Счетчик, "Hedgehog Counter") is built around voice as the primary input method: its own bio states it counts calories and macros from photos, voice and text, but a vc.ru comparison of Russian calorie bots highlights the voice path specifically — natural phrases like "two hundred grams of cottage cheese with honey" get transcribed and understood in context, then logged without extra typing.
Being able to talk instead of type is a real convenience for logging a meal on the go, one-handed, or while cooking, and the bot pairs that with a genuinely minimal visual interface — mostly daily summary reports rather than a dense dashboard.
What isn't confirmed publicly is how it handles poor-quality audio or heavy background noise, which the same voice-first design makes more relevant here than for photo- or text-based competitors. If your environment is usually quiet enough for a clear voice message, this is one of the more frictionless loggers in the category; if not, a text- or photo-first bot may prove more reliable.
Free
| Criterion | Weight | Points | |
|---|---|---|---|
| Functionality | 25 | 7 / 10 | |
| Reliability & speed | 15 | 6 / 10 | |
| Popularity & trust | 15 | 4 / 10 | |
| Price & free tier | 10 | 9 / 10 | |
| Usability | 10 | 8 / 10 | |
| Safety & privacy | 10 | 7 / 10 | |
| Freshness | 5 | 6 / 10 | |
| Localization | 5 | 2 / 10 | |
| No intrusive ads | 5 | 8 / 10 | |
| Score | 100 | 64.0 |
No, photo and text input work too, but voice is the bot's headline feature.
Not confirmed publicly how the bot handles poor audio quality; a clearer recording will likely give a more accurate estimate.