Short answer: You do not need a separate dry-run render to find out what voice will cost - the cost of using voice is surfaced to you before any spend is committed, so you can review the charge and decide whether to proceed without first triggering it.
How cost is shown before you spend
When you work with voice features in the authoring tools, the expected cost is displayed at the point of action - before you confirm the render. This means you can see what a given voice pass will cost and choose to continue or cancel without being charged first.
What this means for your workflow
Because the cost estimate is shown upfront, there is no need to run a separate "dry-run" or test render purely to get a price. Review the displayed cost at the confirmation step, then approve only when you are happy with the amount. If you cancel at that point, no charge is applied.
Edge cases and things to keep in mind
- The upfront cost display covers the voice render you are about to confirm. If your script changes after that point, re-check the cost before re-running voice.
- If you believe the displayed cost does not match what was actually charged, contact support with your project details so the discrepancy can be investigated.