Product overview

What is Picovoice Leopard?

Leopard is Picovoice's on-device speech-to-text engine.

Leopard is Picovoice's on-device speech-to-text engine. It is distributed through SDKs and is intended to be integrated into an application rather than used as a standalone hosted editor.

Who should consider it?

Developers who need embedded or local speech-to-text processing.

Available platform types

Local/CLI

Confirm the exact operating system, device and plan requirements with the provider.

Inside the product

Key features & workflows

These highlights describe the provider’s public offering. Availability can depend on the selected plan or product version.

On-device inference

recognize speech within a supported local environment.

SDKs

access the engine from supported programming environments.

Language models

select the provider's available recognition models.

Make a useful shortlist

What to check in your trial

Check licensing, AccessKey requirements, language-model availability and the hardware supported by the chosen SDK.

Use a fixed dataset representing the language, accents, audio quality and speaker overlap expected in production. Record the exact model and configuration.
Measure recognition quality, latency and failure behavior separately. Test reconnects, retries and incomplete audio in the intended API mode.
Review regions, quotas, retention and request options. Confirm how optional features, channels and long-running jobs affect the bill.

Plans & costs

Picovoice Leopard pricing

The provider’s current pricing has not been reproduced in this profile. Use the official website to confirm available plans, usage limits, billing terms and any trial conditions.

Compare the complete cost

Calculate cost from the actual billing unit and enabled features. Include streaming versus batch modes, channels, minimum charges, optional analysis, storage and network costs where they apply.

Check plans on official site ↗

Explore your options