Free production calculator · text stays on device
Voiceover cost and duration calculator
Estimate spoken runtime and character-based text-to-speech usage before generating a script. Add languages, planned versions, and your provider's current unit price to expose the real size of a voice production job.
What it does
A fast planning estimate
Voiceover budgets grow in three directions: script length, the number of localized versions, and the number of complete generations or revisions. This calculator counts characters and words, estimates spoken duration, and multiplies the text volume by a unit price you provide. It does not claim a vendor price or guarantee final audio length.
Interactive tool
Voiceover cost and duration calculator
Inputs stay in this page. The calculator does not upload a document, image, script, or result.
How to estimate a multilingual voiceover budget
Start with the version of the script that is actually approved for production. A draft with unresolved names, dates, calls to action, or legal language creates a misleading estimate because every later rewrite may require another complete generation. The calculator treats each planned version as a full pass. If your workflow only regenerates individual sentences, use a lower version count and keep a separate contingency for pickups.
For localization, the source character count is only a baseline. German may expand relative to English, Chinese may use fewer written characters while producing a similar spoken duration, and a translator may deliberately rewrite a line to fit a video. Enter the number of target locales to estimate the production surface, then validate each language with its final translated script before purchasing or reserving quota.
Characters measure billing; words help predict runtime
Many text-to-speech services meter usage by characters, credits, or generated seconds. These units answer different questions. Character count is useful for estimating provider usage. Word count divided by narration pace is useful for planning the listener's time. Neither predicts the exact delivery because punctuation, numbers, abbreviations, dialogue, pauses, and pronunciation rules change what the model speaks.
A product name such as “XR-24” may be read as one compact token or expanded into several spoken units. A URL, equation, table, code sample, or legal citation can take much longer than its word count suggests. For a serious production, generate one representative minute with the intended voice and language, measure the result, and use that sample to calibrate the estimate.
Plan revisions without paying for avoidable churn
- Lock product names, numbers, and pronunciation before full generation.
- Split long scripts into stable review sections with clear filenames.
- Review meaning before reviewing voice, pacing, and emotion.
- Ask native reviewers to approve localized scripts before synthesis.
- Keep a version manifest connecting script, language, voice, model, and approval date.
Do not assume the cheapest first pass produces the lowest total cost. A workflow with no approval boundary can regenerate the same material repeatedly. Conversely, do not reserve an arbitrary number of retries when a short pronunciation test would answer the main uncertainty. The useful budget is tied to known review stages.
What this calculator deliberately does not claim
The unit price is editable because provider prices, models, plans, included quotas, commercial rights, and regeneration rules change. Enter the current rate that applies to your account and model. The result excludes taxes, subscriptions, storage, audio editing, translation, human review, music, sound effects, and project management.
EchoVox provides a focused workspace for voice selection, generation history, multilingual production, and developer access. This independent calculator is useful before opening any provider: it makes script volume, localization count, and revision assumptions visible so a team can compare workflows on the same basis.