A medication log app that captures each dose by voice, not memory. Log pills in seconds, keep an accurate record, and walk into appointments prepared.
You take your morning pill, get pulled into a call, and by lunch you genuinely cannot remember whether you swallowed it or just thought about it. A week later a doctor asks how the new dose has been sitting with you, and your answer is a shrug. Paper trackers get left at home. Phone reminders tell you to take a pill but never record that you actually did. Most people managing two or more medications are working from memory, and memory is the worst possible system for this.
A medication log app fixes the record, not just the reminder. The hard part is not remembering to take the pill. It is capturing what you took, when, and how you felt afterward, in the two seconds you actually have. Typing that on a phone keypad while holding a glass of water is why so many logs stay empty.
A reminder and a log are two different jobs, and most apps only do the first one well. A reminder nudges you at 8am. A log answers a question weeks later: did I take the 8am dose on the 14th, and did the headache come before or after I started it?
The second job is where accuracy comes from. Doctors adjust doses based on what you report, and a report built from memory is a guess. A real medication log needs a timestamped entry per dose, room for a short note on how you felt, and a format consistent enough that you can scan three weeks of it and spot a pattern.
The friction is entry speed. If logging a dose takes 40 seconds of tapping, you will skip it on the busy days, which are exactly the days worth recording. The fix is to make the entry as fast as saying it out loud.
Watch how a medication log dies. Week one you fill it in faithfully. Week two you skip the days you are rushed. Week three you stop, because catching up feels like homework. The tracker was never the problem. The keyboard was.
Speaking a dose is faster than typing it, and it works with wet hands, a phone in your pocket, or your eyes on something else. This is where a desktop dictation tool changes the shape of the task. Contextli is a system-level app that types what you say into whatever window is focused, so your spoken note lands directly in your existing notes app, a spreadsheet, or a plain text file. There is no separate medical app to open, no fields to tab through, and no new account tied to your health data.
The distinction matters: Contextli is not a symptom database you upload your life to. It is dictation that produces clean, formatted text wherever you already keep records.
Raw transcription writes down whatever you mumble. A medication log needs structure, and that is the difference a context-aware tool makes.
Contextli uses named writing Modes, and Notes Mode is the one for this. The customization works by example: you feed the Mode 3 to 4 samples of exactly how you want a dose logged, and every future dictation matches that shape. Show it this and it learns the pattern:
2026-07-14, 08:10, Metformin 500mg, taken with breakfast, no side effects noted.
From then on, saying "log metformin five hundred, took it with breakfast, feeling fine" comes out in that same date-drug-dose-note format, cleaned up and ready to paste or already typed into your log. You can add instructions too, like "always write the dose in mg" or "use 24-hour time." That is the customization-by-example flow explained in full here, and it is what turns loose speech into a consistent record you can actually read back later.
This is the "other tools transcribe, we write" idea in practice. The goal is not a wall of transcript. It is a tidy log entry that looks the same every single time.
Medication history is sensitive, and most tracking apps store it on their servers by default. Contextli offers three privacy levels you can stack, which no other dictation tool combines.
First, local models: transcription and the writing step both run on your own machine, so you can turn the internet off entirely and still log a dose. This needs a reasonably modern Mac or Windows laptop; older machines run it slowly, which is the honest trade-off. Second, bring-your-own-key: you supply your own provider API key, and your words go from your machine straight to that provider, so Contextli never sees the audio or text and you pay the provider directly. Third, disable cloud sync: switch it off and Contextli stores nothing in its database, and your notes live as plain local files on your disk.
Stack all three and the whole thing runs offline and private, with your medication record never leaving your laptop. If a company holding your health notes is a dealbreaker, that combination is the point. There is more detail in the dictation privacy guide. One note: a personal log is a helpful record for you and your clinician, not a medical device, so confirm anything clinical with your prescriber.
Here is how spoken logging compares to the common ways people track medication.
| Method | Time per dose entry | Records a timestamp | Format stays consistent | Where the data lives |
|---|---|---|---|---|
| Thumb-typing in a phone notes app | ~40 seconds | Only if you type it | No, drifts every entry | Phone or cloud account |
| Paper pill chart | ~20 seconds | Handwritten, easy to skip | Depends on the printed grid | On paper, easy to lose |
| Reminder-only app | 0, but logs nothing | No, it only nudges | N/A | Nothing recorded |
| Phone dictation into notes | ~15 seconds | Only if you say it | No, raw transcript | Phone or cloud account |
| Contextli Notes Mode on desktop | ~8 to 10 seconds | Yes, in your set format | Yes, taught by example | Your own files, can be fully local |
Phone dictation is faster than typing, but it hands you a raw transcript you still have to reshape. The consistency and the ability to keep everything on your own machine are where a taught Mode pulls ahead. For a wider look at speaking instead of typing, see context-aware speech-to-text.
Consider Dana, who manages three prescriptions and a blood-pressure medication after a recent change. She keeps a plain text file open on her laptop as her running log.
At 8:10am she takes her morning doses, taps the Contextli shortcut, and says: "log the morning three, metformin five hundred, lisinopril ten, atorvastatin twenty, all taken with breakfast, slight headache still there from yesterday." Notes Mode types four clean, timestamped, formatted lines straight into her open file. Total time: about 12 seconds for four doses. Typing that by thumb would have run past two minutes, which is why on a normal week she simply would not have.
When her follow-up appointment arrives three weeks later, she scrolls the file and can see the headache started the day before the new dose, not after it. That is a real answer for her doctor instead of a shrug, and it came from entries that only got made because each one took seconds.
The same routine works for other running records people keep: a food diary alongside a new medication, or a mood tracker when a prescription is meant to affect how you feel. Same dictation, different log.
You do not need a specialized app to start. Pick the place you will actually look at again: a notes app, a spreadsheet, or a text file. Then decide your entry format and teach it to Notes Mode once.
A workable drug log template has five fields per line: date, time, medication name, dose, and a short note. Give the Mode three or four examples in that exact shape, add any rules you want (24-hour time, dose always in mg, one line per drug), and you are done configuring. After that, logging is just talking. The consistency you set up on day one is what makes week three readable instead of a jumble.
Keep the log in one file rather than scattered across apps, and if privacy matters, turn off sync so that file stays on your disk and nowhere else.
It is a tool for recording each dose of medication you take, with the date, time, and often a short note on how you felt. Unlike a reminder, which only nudges you, a log keeps an accurate history you can review later or share with a clinician. Contextli approaches this as a dictation tool that types your spoken entry into whatever log file or notes app you already use.
A reminder tells you to take a pill at a set time. A log records that you actually took it, when, and any effects. Reminders help you not forget; logs give you the history that a doctor can use to adjust a prescription. Many people need both, but the log is the part usually left empty because typing each entry is slow.
Yes. Contextli offers three stackable privacy levels: run the models locally so nothing needs the internet, bring your own provider API key so Contextli never sees your text, or disable cloud sync so your notes stay as local files. Used together, your medication record never leaves your own machine.
You can use your own. Contextli is a system-level desktop app that types into the focused window, so your spoken log entry goes straight into a text file, spreadsheet, or notes app you already keep. There is no separate medical database to sign up for.
A single dose logged by voice through a taught Notes Mode takes roughly 8 to 10 seconds, versus about 40 seconds of thumb-typing on a phone. Logging several doses in one spoken sentence is faster still, which is what keeps the log filled in on busy days.
Yes, if you teach Notes Mode your format first. You give it 3 to 4 examples of exactly how a dose should be logged, plus any rules you want, and every dictation afterward matches that shape. That consistency is what lets you scan weeks of entries and spot a pattern.
No. A personal medication log is a record for you and your prescriber, not a clinical tool or diagnosis. Use it to track and report accurately, and confirm any medical decisions with your doctor or pharmacist.
Contextli is a desktop app for Mac and Windows. Running the models fully on-device for offline, private logging works best on a reasonably modern laptop; older machines will run the local option more slowly.
Build the log your clinician can trust: teach Notes Mode your dose format once and log every pill by voice in seconds, with the option to keep it all on your own machine. Start on the free tier with 100 credits/month, no credit card required, on the Contextli features page, and browse the wider Contexts library for more everyday logging setups.
Junaid Khalid
Founder & CEO
Founder and solopreneur writing about how modern businesses run leaner and faster with AI. I build software that turns everyday work, from capturing thoughts to writing and staying organized, into something effortless, and I share what I learn along the way.
A deposition summary app that types your spoken digest straight into your matter file, keeps witness testimony private, and cuts the hours you spend condensing transcripts.
Junaid KhalidA monthly investor update template for founders, plus a way to draft the note by voice in about 8 minutes instead of an hour of typing.
Junaid KhalidA structured interview debrief template with scoring, evidence prompts, and a fast way to capture notes by voice so your hiring team stays aligned.
Junaid Khalid
