Leopard
Speech-to-Text

cloud-level accuracy, without compromises, and on your terms

Start Building
Talk to Sales

Accurate

Leopard achieves competitive results compared to Watson Speech to Text, Google Speech-to-Text, Amazon Transcribe, and Azure Cognitive Services. Picovoice remains the only ASR vendor that backs its performance claims with open-source and reproducible benchmarks. [1]

Leopard Speech-to-Text accuracy compared to IBM ASR, Google speech recognition, AWS transcribe, and Azure STT
Leopard Speech-to-Text cost compared to IBM speech recognition, Google ASR, AWS transcription, and Azure API

Cost-Effective

10x to 22x more cost-effective compared to cloud-based speech recognition APIs. [1] [2] [3] [4] [5]

Private

All voice data is processed on-device. Intrinsically HIPAA and GDPR compliant.

Zero-Latency

Edge-first architecture eliminates unpredictable network delay.

Reliable

100% of the voice recognition is on-device. No concurrency limit. No downtime.

Free Tier

100 hours of free transcription per month. No credit card is required.

Custom Vocabulary

Understand what matters the most by including terms unique to your use case.

Cross Platform

Runs everywhere. Linux, macOS, Windows, Android, iOS, Raspberry Pi, and NVIDIA Jetson Nano are supported.

Efficient

Breakthrough efficiency built upon proprietary deep learning technology. Runs everywhere even on tiny embedded processors with a compact size of 20 MB. [1]

Leopard Speech-to-Text runtime compared to Mozilla DeepSpeech
Leopard leopard = Leopard.Create("${ACCESS_KEY}");
string text = leopard.ProcessFile("${PATH}");
Build with .NET
leopard = pvleopard.create('${ACCESS_KEY}')
text = leopard.process_file('${PATH}')
Build with Python
const leopard = new Leopard("${ACCESS_KEY}");
const text = leopard.processFile("${PATH}");
Build with NodeJS
Leopard leopard = new Leopard.Builder()
.setAccessKey("${ACCESS_KEY}")
.build();
String text = leopard.processFile("${PATH}");
Build with Java
leopard = Leopard{AccessKey: "${ACCESS_KEY}"}
text, err := leopard.ProcessFile("${PATH}")
Build with Go
Leopard leopard = new Leopard.Builder("${ACCESS_KEY}")
.setModelPath("${MODEL_PATH}")
.build(appContext);
String text = leopard.processFile("${PATH}");
Build with Android
let leopard = Leopard(
accessKey: "${ACCESS_KEY}",
modelPath: "${MODEL_PATH}")
let text = leopard.process_file("${PATH}");
Build with iOS
leopard = await Leopard.create(
"${ACCESS_KEY}",
"${MODEL_PATH}");
String text = = await leopard.processFile("${PATH}");
Build with Flutter
const leopard = await Leopard.create(
"${ACCESS_KEY}",
"${MODEL_PATH}")
const text = await leopard.processFile("${PATH}")
Build with React Native

Need a streaming Speech-to-Text SDK? Try out Cheetah Speech-to-Text.