Credits Usage

The Credits Usage page gives you a transparent log of every credit‑consuming action the web app performs on your behalf, so you can understand which features are using your credits and how much each one costs. Open it whenever you want to verify a charge, audit a spike in usage, or plan ahead.

Opening the Page

The Credits page lives at /credits in the web app.

• Click Credits in the sidebar, or
• Navigate to the route from the account menu

The page loads the most recent usage records by default and lets you filter down to a date range or a specific service.

What the Table Shows

Each row is one usage record — a single call to a paid service.

Column	Description
Date	When the request was processed
Service	The kind of service that was used (TTS, ASR, translation, LLM, assessment, etc.)
Description	A short human‑readable summary of what the request did
Credits	How many credits this call cost
Status	Whether the call succeeded, failed, or was refunded

The list is paginated 50 records at a time. Use the Previous and Next buttons to move through history, and Clear Filters to reset your view.

Filtering Usage

You can narrow the table to find specific records.

By Date Range

Use the Start Date and End Date inputs to bound the period you care about. The filters accept dates in YYYY-MM-DD format. Changing either date resets pagination to the first page.

By Service Type

The Service Type selector lets you pick one of the following categories:

• All — every service combined
• TTS — text‑to‑speech synthesis
• ASR — automatic speech recognition (transcription)
• Translation — machine translation
• LLM — large language model completions
• Assessment — pronunciation and speaking assessments

This is the fastest way to answer questions like “How many credits did I spend on translation last week?”

Reading a Record

A typical record looks like:

Field	Example
Date	Jun 09, 2026
Time	10:14:32
Service	TTS
Description	Generate speech for the word "bonjour"
Credits	1
Status	Succeeded

The description is intentionally short — it is meant to remind you of the request, not to reproduce it. For full context, return to the feature that made the call (for example, the transcript or chat thread that triggered it).

When a Charge Looks Wrong

If a row does not match what you did, try these checks before contacting support:

1. Check the service and description — many features share a service (for example, dictionary lookups use LLM credits). Make sure the row corresponds to the feature you remember using.
2. Check the timestamp — the timestamp reflects when the request was processed, not when you clicked. Background jobs (sync, transcription, AI) may run minutes after the originating action.
3. Filter to a tighter window — re‑filter the table to a single day and service to see only the rows that match the charge you are questioning.
4. Look at status — failed requests are usually refunded. If the row says Failed, the credits may already be back on your balance.

Tips for Managing Usage

• Run heavy jobs when you can wait — large transcriptions and bulk translations are the biggest single charges. Schedule them for times when you can let them finish uninterrupted.
• Reuse cache — many services (dictionary lookups, repeated translations) cache results, so re‑running the same task the second time is usually free.
• Watch the LLM bucket — chat and free‑form AI features are flexible but easy to over‑use. Review this page weekly if you are on a tight plan.
• Export your history — if you need an archive of your usage for accounting, take a screenshot of the filtered table or copy the rows you need before clearing the filter.

Limitations

• Only requests made from the web app appear here. The browser extension has its own usage trail, viewable from the extension’s settings.
• Records are kept according to the retention policy of the backend. Very old entries may no longer be downloadable.
• The page shows individual records, not aggregate charts. To see trends, export the rows you need and chart them in a spreadsheet.