Turn calls, meetings and uploaded audio into task-ready transcripts.
Speaker diarizationAvailable
Model · Whisper Large V3
50 creditsper started audio second
≈ $0.00005 / second
Each started audio second is billed individually.
Usage-based pricing
Pay for the tasks you run.
Use one credit balance across task-specific AI capabilities. Each capability keeps a rate and usage unit that matches the work.
Credit exchange
$1
USD
1M credits
shared task balance
Add credits when needed and pay only for measured usage.
Task rates
A clear unit for every task.
Use the same credit balance while each capability is measured in the unit that best represents its work.
Generate natural speech from text for products and agent workflows.
Voice cloningComing soon
Pricing will follow the task's natural usage unit.
Character-based pricing will be published at launch.
Separate speech, music and other sources for downstream audio tasks.
Source separationAvailable
Model · Managed separation
20,000 creditsper started minute
≈ $0.02 / minute
Each started audio minute is billed as one whole minute.
One balance, many tasks
Billing follows the work.
Credits stay consistent across TaskInfer while metering follows the natural unit of each capability.
- 01One credit balanceAdd credits once and use the balance across available tasks.
- 02Task-specific ratesEach capability publishes a rate in its natural usage unit.
- 03Measured consumptionTask usage is translated into credits at the published rate.
Start with an available capability
Put specialized inference to work.
Create an API key and put Speech to Text or Audio Separation to work with shared Credits and task-specific metering.