> ## Documentation Index
> Fetch the complete documentation index at: https://api-docs.voicecheap.ai/docs/llms.txt
> Use this file to discover all available pages before exploring further.

# Create a project

> Set up a new VoiceCheap project before transcription, translation, and dubbing begin.

# Create a project

Create a project when you want VoiceCheap to ingest your source media, prepare the transcription, and set the language and context for the rest of the workflow.

## Before you start

* Choose the source media you want to translate or transcribe
* Know the target language you want to deliver
* Decide whether you want to leave the original language on auto-detect or set it yourself
* Gather any saved brand vocabulary or subtitle files you want to reuse

## What you decide before transcription

| Setting           | Why it matters                                                                                     |
| ----------------- | -------------------------------------------------------------------------------------------------- |
| Input source      | Defines what VoiceCheap should ingest and validate first                                           |
| Target language   | Defines the translated delivery language for the project                                           |
| Original language | Helps when you know the source language with certainty, but auto-detect is often the safest choice |
| Project name      | Makes project history and collaboration easier to manage                                           |
| Speaker count     | Improves how the transcript is segmented by speaker                                                |
| Brand vocabulary  | Helps the transcription phase recognize names, brands, and uncommon terms                          |
| Existing SRT      | Lets you start from a subtitle file when you already have one                                      |

## Supported source types

* Local video or audio uploads
* YouTube imports
* TikTok imports
* Instagram imports
* X/Twitter imports
* Google Drive files
* Public links to supported video or audio files

VoiceCheap validates each source type before the project is created. If a direct link is public but points to an unsupported file type or an inaccessible file, import can fail before transcription starts.

## Project creation flow

<Steps>
  <Step title="Choose an input source">
    Upload a local file or import supported media from a link-based source. VoiceCheap validates the source before the project is created.
  </Step>

  <Step title="Set the language pair">
    Select the target language first. If you already know the original language and want to lock it in, you can set that too. Otherwise, leave source detection on automatic.
  </Step>

  <Step title="Add project structure">
    Give the project a recognizable name and set the number of speakers when you want better diarization and transcript structure.
  </Step>

  <Step title="Add transcription context">
    Use your saved brand vocabulary for names, product terms, or domain-specific wording. If you already have subtitles, you can also provide
    an SRT file.
  </Step>

  <Step title="Start transcription">
    Once the project starts, VoiceCheap begins the transcription phase and prepares the project page for the later translation and dubbing steps.
  </Step>
</Steps>

## Brand vocabulary vs glossary

| Feature          | Scope                                        | What it helps with                                                       |
| ---------------- | -------------------------------------------- | ------------------------------------------------------------------------ |
| Brand vocabulary | Saved account or team customization          | Improves transcription for names, products, and uncommon terms           |
| Glossary         | Translation customization by source language | Preserves or forces translation wording during translation and SmartSync |

Brand vocabulary is not meant to control translation wording. It improves transcription by giving the transcription model a better chance to recognize uncommon terms before the dubbing workflow even starts.

Good practice:

* add product names, brand names, people names, acronyms, and uncommon terms
* keep entries concise
* separate entries with commas
* stay under the current maximum of 100 entries

Example:

`VoiceCheap, Raycast, voicecheap.ai, chat-gpt, Claude, Anthropic, dubbing, timestamps, Firecrawl, Claude 4 Opus, OpenAI, Cursor`

## What happens next

After creation:

* VoiceCheap validates the upload or import source and creates the project
* Transcription starts first, with fallback behavior across supported providers when needed
* The project page becomes your review area once the source transcript is ready
* The best next step is usually to open [Dubbing settings](/docs/dubbing/dubbing-settings) before you launch the translation
* Translation, voice selection, subtitles, lip sync, and publishing happen later in the workflow

## Related pages

* [Customize translation projects](/docs/dubbing/customize-a-project)
* [Dubbing settings](/docs/dubbing/dubbing-settings)
* [Brand vocabulary](/docs/customization/brand-vocabulary)
* [Glossary](/docs/customization/glossary)
* [How VoiceCheap works](/docs/documentation/how-voicecheap-works)