Question 1

What can I create with VocalCopyCat?

Accepted Answer

Create voiceovers for stories, videos, lessons and other audio projects. You can generate speech from text, use a reference recording for voice cloning, save reusable voices, and edit projects by paragraph. Developers and compatible AI tools can use the same account through the REST API and MCP.

Question 2

Can I try it for free?

Accepted Answer

You can try a short homepage preview without an account. New accounts receive 2,000 welcome credits. The current packages are one-time purchases of prepaid credits; buy more when you need them. Saving a reusable custom voice costs 10,000 credits, so the welcome balance alone does not cover that action.

Question 3

Which credit package should I choose?

Accepted Answer

Starter: $5 for 100,000 credits; Creator: $35 for 1,750,000 credits; Premium: $100 for 10,000,000 credits. All are one-time purchases with the same tools. Premium has the best package rate. Custom top-ups start at $10 for 300,000 credits; the rate rises smoothly from 30,000 credits per dollar at $10 to 100,000 at $100, then stays at 100,000 per dollar. The maximum purchase is $10,000. Review the quoted credit total before checkout.

Question 4

What do the package savings percentages compare?

Accepted Answer

The package cards compare the cost of library speech with ElevenCreative Starter monthly website text-to-speech using Multilingual v2 at its base credit rate. Open the savings details beside the packages for the checked date, official sources and calculation assumptions. The figures assume the full package and monthly allowance are used for normalized English ASCII text. They do not apply to custom/reference speech, separate ElevenLabs API rates, discounts or unused credits, and do not claim equivalent quality or features.

Question 5

How do I generate and download speech?

Accepted Answer

Open Create speech in your workspace, choose a library or saved voice, enter your text and check the credit estimate. Submit once and follow the job's progress. When it completes, listen and download the result. Signed-in generations are available in Recent generations, including those created through the API or MCP. Each generation accepts up to 5,000 input characters; the public text preview is limited to 120.

Question 6

How does cloning differ from saving a voice?

Accepted Answer

One-off cloning uses a reference recording for that generation. My voices saves a reference and transcript so you can select it again in the workspace or integrations. Saving a voice costs 10,000 credits and does not train a separate model. Choose a clear audio file up to 15 MB that you own or have permission to use, then trim and preview a section under 59 seconds. Recordings use the same clip editor. Trimming happens on your device; only the prepared clip is uploaded.

Question 7

How are credits calculated?

Accepted Answer

Library speech costs one credit per UTF-8 byte after text normalization. Custom voices and one-off cloning cost five credits per byte; saving a reusable voice costs 10,000 credits. Some characters use several bytes, so character count and cost can differ. The workspace, REST API and MCP use your shared account balance. Accepted generation jobs reserve credits; a terminal failure or cancellation while still queued returns the reservation once.

Question 8

Can I use VocalCopyCat through an API?

Accepted Answer

Yes. Open Developers in your workspace to create a key with selected permissions, an expiry and a cumulative credit budget. The REST API can list available voices, estimate cost, create speech or one-off clone jobs, check results and usage, and cancel work that is still queued. Keep the secret on your server; it is shown only once. There is no separate API credit balance.

Question 9

Can my AI assistant use VocalCopyCat through MCP?

Accepted Answer

A compatible remote MCP client can connect to https://www.vocalcopycat.com/mcp. Review the requested permissions and credit budget before approving access. The five tools list voices, estimate a generation, submit speech, check a job and read your balance. OAuth access lasts one hour and requires authorization again after expiry; a client that supports bearer credentials can instead use a scoped developer key. Client support varies.

Question 10

Can I control or revoke integration spending?

Accepted Answer

For a developer key, choose permissions, expiry and a credit budget. For an OAuth connection, review the requested permissions and set its budget; access expires after one hour. A budget limits cumulative spent and currently reserved credits; it is not a monthly allowance and does not add credits to your account. You can revoke access in Developers. Revocation blocks future credential use; it does not undo jobs already accepted.

Question 11

Why is my generation waiting?

Accepted Answer

Generation jobs share limited processing capacity and may wait for earlier work or model startup. Follow the job status instead of submitting repeatedly. If your connection drops, keep the existing job and reconnect to it. API clients should reuse the same idempotency key and input after an uncertain submission, poll only while a job is active, and stop polling at a terminal status.

Question 12

Can I cancel a generation?

Accepted Answer

You can cancel a job while it is still queued, which returns its reserved credits. Work that has already started cannot currently be cancelled. The REST API supports queued cancellation; there is not currently a dedicated MCP cancellation tool.

Question 13

How do projects and paragraph editing work?

Accepted Answer

Use Projects to keep a longer script in an addressable project. Save paragraph edits, generate their audio, then export when every paragraph has current saved audio and no unsaved changes. Editing a paragraph makes its old audio stale until you generate it again. A conflicting save is reported so you can review it rather than overwrite a newer version.

Question 14

Are my drafts saved across devices?

Accepted Answer

Unsubmitted editor drafts are saved locally in the current browser when storage is available. Explicitly saved projects and completed signed-in generations belong to your account. Local drafts are not a cross-device backup and may disappear if browser data is cleared. Recent generations shows your latest 30 jobs.

Question 15

Why does a download link expire?

Accepted Answer

Saved customer audio is stored privately. Its signed storage link lasts 15 minutes; that is the link's lifetime, not a promise that the audio is deleted after 15 minutes. Reopen the saved generation, or use an authorized integration to request a fresh result link. Treat a signed link as private because anyone holding it can use it until it expires.

Question 16

Can I create a reusable voice through MCP or the API?

Accepted Answer

Create reusable voices in My voices in the browser. Once saved, integrations can select them from the available voice list. The REST API also accepts a reference recording for a one-off clone. MCP currently submits speech using available voice IDs and does not upload audio or create saved voices.

Question 17

Which languages and accents are supported?

Accepted Answer

Language and accent quality depend on the voice model and your reference audio. Listen to samples and use a short preview to test your intended language. Clear references and accurate transcripts help; review the generated speech before publishing it.

Question 18

Does the API stream speech or send completion webhooks?

Accepted Answer

The current API and MCP return asynchronous jobs and completed audio links. Streaming speech, batch submission, completion webhooks and OAuth refresh tokens are not currently provided. Poll an active job at a bounded interval and stop when it completes, fails or is cancelled.

Question 19

What do the playback controls change?

Accepted Answer

Playback controls let you listen, pause, seek and adjust how you hear an existing recording. They do not regenerate the voice, change the saved audio file or spend generation credits. Starting another preview pauses the previous one.

Question 20

What recordings should I use for cloning?

Accepted Answer

Use recordings you own or have permission to use. Choose clear speech from one speaker with little background noise, and check the transcript when saving a voice. Generated results vary with the reference and text. Read the service terms before publishing or sharing your output.

Questions, answered.

A few things
you might wonder.

A few thingsyou might wonder.

01What can I create with VocalCopyCat?

02Can I try it for free?

03Which credit package should I choose?

04What do the package savings percentages compare?

05How do I generate and download speech?

06How does cloning differ from saving a voice?

07How are credits calculated?

08Can I use VocalCopyCat through an API?

09Can my AI assistant use VocalCopyCat through MCP?

10Can I control or revoke integration spending?

11Why is my generation waiting?

12Can I cancel a generation?

13How do projects and paragraph editing work?

14Are my drafts saved across devices?

15Why does a download link expire?

16Can I create a reusable voice through MCP or the API?

17Which languages and accents are supported?

18Does the API stream speech or send completion webhooks?

19What do the playback controls change?

20What recordings should I use for cloning?

A few things
you might wonder.