Energy and battery life: what the dictation models cost your Mac
By Ryleigh Newman · Published · Updated
Between dictations, YapToText lets go of the microphone about a second after each one and, by default, unloads its models two minutes after use, so what it costs your battery comes down to three settings you control. Two of them are on the Energy page in the sidebar: how long the models stay in memory, and whether lighter models run on battery. The third, Keep the microphone warm, is in Settings, General, Microphone.
Everything runs on the Mac itself, so the only question is how much of your Mac the app may use, and when. It matters most on a MacBook running on battery.
This app is completely free. You don't have to make an account. There are no subscriptions, no difficult downloads, and nothing is locked. Made for Apple silicon Macs. Source on GitHub.
On this page: 11 sections
Steps
- Open the Energy page in YapToText and check the Power line in the This Mac card.
- Turn on Switch models with the power source in the Adaptive models card.
- Under Cleanup model, choose a lighter model in the On battery menu. This is where the biggest saving is.
- Under Dictation model, choose a lighter model in the On battery menu, or leave it on Use the main selection.
- In the Memory card, set Keep models in memory. Only while dictating saves the most, and Until quit answers quickest.
- In Settings, General, Microphone, leave Keep the microphone warm off for the longest battery life.
Where it is
The Energy page has three cards:
- This Mac. The chip, memory, core count, model identifier and power state, with the battery percentage on a laptop. A desktop Mac has no battery, so it always reads as plugged in.
- Memory. Keep models in memory, the setting that decides how long the models stay loaded.
- Adaptive models. Switch models with the power source, and the plugged-in and on-battery choices it opens.
Keep the microphone warm sits with the other microphone settings in Settings, General, Microphone. The Statistics page adds an Energy & resources readout of memory now, CPU now and the model unload delay, so you can watch the cost on your own Mac. By default the app lets go of the microphone and the models when you are not dictating.
Run lighter models on battery
Switch models with the power source is off by default. Turn it on in the Adaptive models card and the app runs different models depending on whether the Mac is plugged in. The cleanup model is the heaviest thing the app runs, so a lighter one on battery saves the most.
Step 1: Turn it on
- You: On the Energy page, turn on Switch models with the power source.
- YapToText: Shows two groups, Dictation model and Cleanup model, each with a Plugged in menu and an On battery menu.
- Check: All four menus start on Use the main selection, which keeps whatever the AI Models page has chosen.
Step 2: Pick the battery models
- You: In the On battery menu under Cleanup model, choose a lighter model, and do the same under Dictation model if you like.
- YapToText: Uses those models whenever the Mac runs on battery, and the plugged-in choices when it is on power.
- Check: The model you picked is downloaded on the AI Models page. The menus list every model in the library, your own included and marked (yours), and a model that is not on disk is skipped for another installed one.
The AI Models page shows each model's size with star ratings for accuracy and speed. In the app's own words, bigger models hear better, and smaller ones are kinder to your battery. A mode can also carry its own models in its editor, including an on-battery speech model, and a mode's choice wins over these defaults.
Choose how long the models stay in memory
A loaded model answers right away but holds memory the whole time it sits there. Keep models in memory, in the Memory card, decides how long the speech and cleanup models stay loaded after a dictation. Unloading sooner saves energy and memory, and the next dictation after an unload takes a few extra seconds.
The choices are Only while dictating, 10 seconds, 30 seconds, 2 minutes or 15 minutes after use, and Until quit. The default is 2 minutes after use. Only while dictating is the deepest saver. Until quit is the quickest to answer and holds the memory for as long as the app is open.
While a model reloads, the panel says Loading the speech model, or Loading the AI model when it is the cleanup model, so a slow stop does not look frozen. The speech model also preloads a few seconds after launch, and so does the cleanup model when Post-transcription analysis is on. After an unload, the speech model warms again in the background when you use the keyboard or mouse, and the same timer then lets it go again.
When macOS runs short of memory, YapToText drops its loaded models at once and waits five minutes before warming them again.
Keep the microphone warm, or let it go
Keep the microphone warm is in Settings, General, Microphone, and it is off by default. Off, the microphone lets go about a second after each dictation, and the orange indicator shows only while you dictate. The app's own note says the next dictation may then miss its first second.
On, the input stays open after a dictation, and the app's note says it is ready instantly for the next one. A Stay warm for slider sets how long, from 1 to 30 minutes, starting at 5. When that window runs out, the microphone lets go, the same as if the setting were off. A warm microphone costs an open input and an indicator that stays lit.
With the setting off, launching the app or switching apps never opens the microphone on its own.
Other things that cost battery
- Live transcription preview. On by default in Settings, Advanced, it shows your words while you speak, and the app notes that it costs extra battery with Whisper models. Turn it off if you never watch the panel.
- A heavier model. A larger speech or cleanup model hears or writes better and costs more battery and memory on every dictation.
- A dictation itself. The app's readout notes that CPU spikes briefly during transcription and cleanup, then returns to about zero at idle. The CPU now figure on the Statistics page lets you check that on your Mac.
What I would pick
If I wanted the longest battery life, I would put a lighter cleanup model on the battery side, set Keep models in memory to Only while dictating and leave the microphone cold. If I wanted the quickest dictation, I would do the opposite: Until quit and a warm microphone. The defaults sit in between: the same models on both power sources, a 2 minute unload timer and a cold microphone.
Limits
- Desktop Macs. With no battery, a Mac always reads as plugged in, so the on-battery choices never apply. The unload timer and the warm microphone work the same everywhere.
- No battery numbers. This page gives none. The Energy & resources readout on the Statistics page shows memory and CPU on your own Mac.
- Models you have not downloaded. Picking one for battery does nothing until it is on disk. Another installed model stands in.
If something is off
- The first dictation after a break is slow. The models unloaded and are reloading. Choose a longer Keep models in memory time, or Until quit.
- The orange microphone indicator stays on. Keep the microphone warm is on. Turn it off in Settings, General, Microphone, or shorten Stay warm for.
- The battery model is not used. Check that Switch models with the power source is on, that the This Mac card reads On battery, that the model is downloaded, and that the mode you dictate in does not carry its own model.
- Your Mac feels short of memory. Pick Only while dictating or a short timer, and choose smaller models. When macOS runs low, the app drops its models by itself.
If dictation itself misbehaves, the troubleshooting page goes step by step.
The short version
- For battery. Use a lighter cleanup model on battery, Only while dictating, and a cold microphone.
- For speed. Choose Until quit and a warm microphone.
- For most people. Keep the defaults: 2 minutes, the same models everywhere, and a cold microphone.
Three settings decide what YapToText costs your battery, and the defaults are a fair middle. If your Mac behaves differently from this page, tell me on GitHub. Every issue gets read.
This app is completely free. You don't have to make an account. There are no subscriptions, no difficult downloads, and nothing is locked. Made for Apple silicon Macs. Source on GitHub.
Related
Questions
Does YapToText use battery when I am not dictating?
By default it lets go of the microphone about a second after each dictation and unloads its models two minutes after use. Keep the microphone warm is the one setting that keeps the input open between dictations, and it is off by default.
Why is my first dictation after a break slower?
The models unloaded after the Keep models in memory time and have to reload, which takes a few extra seconds, and the panel says so while it happens. Choose a longer time on the Energy page, or Until quit, to skip the wait.
Why does the orange microphone indicator stay on after I stop?
Keep the microphone warm is on, so the input stays open for the minutes set on the Stay warm for slider. Turn it off in Settings, General, Microphone, and the microphone lets go about a second after each dictation.
Can I run a smaller model only on battery?
Yes. Turn on Switch models with the power source on the Energy page and pick a lighter model in the On battery menu for the dictation model, the cleanup model or both. Leave Plugged in on Use the main selection to keep your usual choice there.
Does the Energy page do anything on a desktop Mac?
It still shows the chip, memory, cores and model identifier, but a Mac with no battery always reads as plugged in, so the on-battery choices never apply. Keep models in memory and Keep the microphone warm work the same on any Mac.