Voice Input and AI Task Creation

Voice Input and AI Task Creation

Voice input helps capture a task quickly when the details are easier to say than type on a phone. Open a new or existing task form: the microphone button sits to the right of the Primary information heading. The feature works on devices where iTasks has permission to use the microphone.

1. Complete flow

The working sequence is free-form dictation → automatic recognition and AI structuring → ready task → optional fine-tuning of the filled fields → save. You do not need to proofread or edit the raw dictation first.

1. Open the task form and tap the microphone. Allow microphone access if the device asks for it the first time.
2. Describe what should be done, the expected outcome, and details that must not be lost. The panel lets you pause, resume, or stop recording.
3. Stop recording and press Add to task. AI interprets natural speech, removes filler, identifies the important points, and prepares the title and description.
4. If the title or description already contains text, the button reads Update task. The new dictation is merged with the current fields instead of overwriting them.
5. If needed, fine-tune the ready task, then manually choose the project, assignee, due date, type, priority, sprint, and any other fields you need.
6. Create or save the task with the form's regular action.

2. How to choose the task's output language

By default, AI writes the title and description in the language of the original dictation. To create the task in a different language, name the target language while recording. For example, dictate the details in English and add: “Prepare the final task in Persian.” AI removes this meta-instruction from the content and returns the title and description in Persian.

You can ask for Russian, English, Arabic, Spanish, or any other language in the same way. Before saving, review professional terms, names, and the expected outcome, especially when the output language differs from the language you spoke.

3. How to add to or correct a task by voice

You do not need to erase the existing content or start over. Open the filled form, tap the microphone again, and say what should be added, clarified, or corrected. For example: “Add an acceptance criterion: test on iOS and Android” or “Change the date in the description to August 20.”

When you press Update task, iTasks sends the new dictation to AI together with the current title and description. AI must return complete updated versions and preserve important details from the previous content unless the dictation explicitly asks to remove or replace them. This lets you build a task over several voice additions.

4. What to verify before saving

AI helps structure what you said, but it does not verify facts or make the final decision for you. Check names, dates, amounts, links, expected outcomes, and acceptance criteria. Also verify the fields AI does not fill: project, assignee, due date, type, priority, watchers, workflow, and recurrence.

Review the prepared task rather than the raw transcript. If you explicitly asked AI to remove or replace earlier content, make sure the result matches that instruction.

5. What AI tokens mean

AI tokens are an internal allowance for speech recognition and task drafting. They are not Telegram Stars and not a PRO subscription. Recognition and AI drafting are counted as separate operations, so usage depends on the recording length and processing required.

All workspace members can use built-in AI features and draw from the shared workspace allowance. Voice task input is currently the only built-in AI feature. The allowance is not split into personal packages: every member's operations reduce the same available balance.

The owner sees the balances on the subscription page in the AI tokens row. Pressing it opens the usage report: Remaining (subscription) applies only to the current Free or PRO periodic allowance, while Remaining in packs is the sum of the remaining balances in all non-depleted packs. The report also breaks usage down by user and by feature. Regular members can use the shared workspace resource, but purchases and the billing report are owner-only.

When the balance reaches zero, a new dictation cannot start. iTasks shows the allowance renewal date and the action available for the current plan. Existing tasks and regular manual editing are not removed.

6. Purchase, confirmation, and refunds

1. Select the pack on the Subscription screen, accept the purchase terms, and open the Telegram Stars invoice. Tokens are granted only after successful payment and server confirmation.
2. If you closed or cancelled the invoice before payment, no Stars are charged and no pack is granted. You can start the purchase again.
3. If payment succeeds but confirmation from Telegram is delayed, do not pay a second invoice. Wait a few seconds and refresh the status. If the pack still does not appear, use /paysupport or contact support.
4. An approved refund is completed in Telegram Stars and marked in purchase history. A refund does not revoke AI tokens that were already granted: the remaining balance stays with the same workspace and continues to follow the normal consumption rules. If a payment error prevented delivery, no tokens are added to the balance.

7. Open beta and feedback

All built-in AI features are currently in open beta. The available functionality, usage rules, AI token allowances, and other information shown on this page or in the app may change in the future based on testing results.

Send suggestions for improving voice input or other iTasks features to the support bot @itaskssupportbot.

8. Privacy and user control

During dictation, the audio stream is sent to the speech recognition service. The recognized text is processed by an AI service to prepare a task title and description. Do not dictate secrets, passwords, payment credentials, or unnecessary personal data. The current data categories, purposes, and providers are described in the Privacy Policy.

The recognized text is visible as process feedback, but you do not need to edit it first: AI structures the dictation automatically. After applying it, the fields remain editable, and the task is created only when you use the form's regular save action.

9. Troubleshooting

• No microphone access — allow it for Telegram, the browser, or the app in device settings and try again.
• The connection drops — stop recording, check the network, and start again. Review the filled task fields after AI processing.
• The allowance is exhausted — open the subscription screen to check the balance and renewal date.
• A paid pack is missing — refresh the status after a few seconds; do not create a duplicate payment while the first one is being checked.
• The workspace is Read only — restore working access first; AI tokens do not bypass plan restrictions.

Related guides

All task fields are covered in Creating and Editing a Task. Plans, limits, and Read only behavior are covered in Subscription, Plans, and Limits. The product overview is on AI Voice Task Creation. You can also read the Terms of Use, the Privacy Policy, use /paysupport for payment questions, or contact support@itasksbot.com.