Voice and text use natural-language understanding, while manual entry lets you fill in fields directly without AI parsing. Always check the units and the result.
Speak naturally, but make the essentials explicit: exercise name, reps, load or duration, then optional RIR, tempo or set type. For circuits, also mention duration or number of rounds.
Yes. Separate movements with words such as 'then' or 'next', or with clear sentences. Say when you change movements so sets attach to the right exercise.
No. When a workout is in progress, 'another 10 at 40 kg' can be added to the most recent exercise. In one dictation, 'then another 3 at 110' applies to the exercise just named.
You can correct yourself out loud, for example 'sorry, 75 kg, not 70'. Still, never assume the correction was understood: verify the load and reps in review before saving.
Equipment helps identify the right catalog variant. A dumbbell bench press is not the same exercise identity as a barbell bench press in your history. Use consistent variants for meaningful comparisons.
At the exercise clarification step, use search to select the movement you actually performed. Do not pick an unrelated exercise just to dismiss the warning.
If a spoken exercise is misidentified, choose the correct catalog variation in Review workout. In recent versions, Grunty can learn that phrase as your personal alias so future uses can find the right movement without changing its official name for everyone.
Speak close to the microphone, use shorter phrases and review the result. If the environment is too noisy, natural-language text is a practical alternative.
Sending and interpreting dictation requires connectivity. Without a network, Grunty shows a readable error and asks you to dictate again. Do not assume the session was saved until you see confirmation.
When listening is enabled on a compatible device, 'Hey Grunty' can start voice entry. You can also use the microphone button, especially in a noisy environment.